Split the sequence implementation into smaller stages

AGENTS.md

@@ -1,35 +1,25 @@
-Please review `docs/development.md` for initial orientation in this repository
-and follow its task-specific reading guide.
+<!-- codebase-memory-mcp:start -->
+# Codebase Knowledge Graph (codebase-memory-mcp)
 
-## Implementation Discipline
+This project uses codebase-memory-mcp to maintain a knowledge graph of the codebase.
+ALWAYS prefer MCP graph tools over grep/glob/file-search for code discovery.
 
-Simplify the mechanism, not the user's choices.
+## Priority Order
+1. `search_graph` — find functions, classes, routes, variables by pattern
+2. `trace_path` — trace who calls a function or what it calls
+3. `get_code_snippet` — read specific function/class source code
+4. `query_graph` — run Cypher queries for complex patterns
+5. `get_architecture` — high-level project summary
 
-**Keep implementation machinery proportionate.** Use straightforward,
-idiomatic code and introduce only the machinery needed to support the requested
-behavior and flexibility. Optimize for conceptual simplicity and
-maintainability, not for the fewest lines, fewest options, or narrowest product
-surface.
+## When to fall back to grep/glob
+- Searching for string literals, error messages, config values
+- Searching non-code files (Dockerfiles, shell scripts, configs)
+- When MCP tools return insufficient results
+<!-- codebase-memory-mcp:end -->
 
-**Do not invent restrictions.** Do not reject, prohibit, constrain, cap,
-harden, or otherwise narrow behavior unless required by the user, an explicit
-repository contract, or an unavoidable correctness constraint. Preserve
-operator and caller choice where policy has not been established. If an
-additional restriction appears advisable, present it as a decision rather than
-silently implementing it.
+Start with `docs/development.md` for orientation and use its policy reading guide
+to select the policies applicable to the task. Do not read the entire suite by
+default.
 
-A correctness constraint must follow from the requested behavior or an
-established invariant, not from the convenience of the chosen implementation.
-Explain the specific failure it prevents and why a less restrictive
-implementation cannot satisfy the contract.
-
-**An implementation detail is not repository policy.** Tests and documentation
-must not elevate incidental behavior into a durable guarantee or restriction
-unless that behavior was expressly selected as part of the contract. Existing
-explicit contracts remain authoritative; changing or weakening one requires a
-deliberate decision rather than being inferred from this principle.
-
-Agent-authored plans, tests, documentation, and prior implementations are not
-evidence that an additional restriction was deliberately approved. Identify the
-authorizing requirement or decision; when that authority is unclear, surface
-the ambiguity rather than treating repetition as approval.
+Read and follow `docs/policy/agents.md` before recommending, planning, or
+implementing repository changes.

COPYRIGHT

@@ -0,0 +1,25 @@
+ConfigKit
+
+Copyright (C) 2026 Eric Rakestraw
+
+SPDX-License-Identifier: LGPL-3.0-or-later
+
+This library is free software: you can redistribute it and/or modify it
+under the terms of the GNU Lesser General Public License as published by
+the Free Software Foundation, either version 3 of the License, or
+(at your option) any later version.
+
+The GNU Lesser General Public License version 3 and its incorporated
+GNU General Public License version 3 text are provided in LICENSE.
+
+This library is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+GNU Lesser General Public License for more details.
+
+Third-party code and assets are subject to their respective license terms.
+
+Policy documents adapted from Standard Repository Policies, the root AGENTS.md
+orientation, and the Policy Reading Guide in docs/development.md remain licensed
+under CC-BY-SA-4.0. See docs/oss/attribution.md for the affected material,
+upstream credit, and license links.

README.md

@@ -5,18 +5,42 @@ An opinionated Go library for applications that use YAML configuration files.
 Linux is release-validated; macOS support is best-effort. See the
 [platform support policy](docs/policy/release.md#platform-support) for details.
 
-ConfigKit's initial loader accepts one strict YAML
-source (or explicit defaults-only input), optionally traverses explicit path and glob
-imports, strictly composes their disjoint definitions, applies typed defaults, and
-can overlay one selected fixed profile and check an application schema version. It
-returns detached metadata and safe diagnostics.
+ConfigKit's simple `Load` accepts one strict YAML source (or explicit
+defaults-only input). `LoadLayered` additionally combines explicit lower-priority
+source groups with an independently selected main input. Sources can be host
+files, native `fs.FS` files, labeled bytes or readers, or snapshotted mappings.
+Enabled explicit imports compose strictly within their group; precedence
+between groups combines mappings recursively. Both entry points return a complete,
+validated configuration with detached metadata and safe diagnostics.
 Applications may perform audited defaulting and normalization through a resolver,
 then report domain errors and warnings through read-only semantic validation. Secret
 tags and a path predicate drive deterministic redacted effective YAML. Typed secret
 references resolve environment names before validation, and explicit
 secrets-directory bootstrap is available as a separate startup operation. Explicit
-CLI override records apply above profiles. Value explanation is available through
-detached structured records and deterministic redacted output.
+CLI override records apply above overlays. Value explanation is available through
+detached structured records and deterministic redacted output. Built-in
+`extends` relationships let a mapping inherit from a named sibling during the
+same load, retaining explicit presence and source/derivation history. The root
+`Diff` API compares frozen effective values from two successful same-type loads;
+the [configuration contract](docs/config.md#effective-configuration-diffs) and
+[consumer guide](docs/consumers/README.md#effective-configuration-diffs) describe
+its safe changed-region records and maintained workflows.
+
+Sources can also be explicitly placed as sequence items, and definitions can
+register selected sequences to append records with distinct string IDs during
+strict composition. Higher-priority sequences still replace lower-priority
+sequences as whole values. See the [sequence contract](docs/config.md#strict-document-composition),
+[maintained prompt example](docs/consumers/README.md#identified-prompt-sequences),
+and [ADR 0014](docs/adr/0014-sequence-placement-and-composition.md).
+
+Imports are also explicit: a source may name files, search declared directories,
+or select files with path globs. Applications can set load-wide import defaults
+while keeping declaration-specific exceptions. Files are acquired in authored
+declaration order; dependency-ready composition uses a shared stable scheduler.
+ConfigKit does not search conventional application paths implicitly. See the
+[import contract](docs/config.md#explicit-imports), [consumer examples](docs/consumers/README.md),
+and [ADR 0015](docs/adr/0015-shared-dependency-ready-ordering.md) and
+[ADR 0016](docs/adr/0016-independent-import-controls.md).
 
 ```go
 type Config struct {
@@ -33,10 +57,48 @@ func ReadConfig(path string) (Config, error) {
 }
 ```
 
-The [configuration contract](docs/config.md) describes implemented behavior. The
-[consumer adoption guide](docs/consumers/README.md) provides maintained Narratio-,
-Notarius-, WeatherReporter-, and Seriatim-shaped examples and migration guidance.
-The [deferred scope](docs/roadmap/deferred.md) records deliberate exclusions. See the
-[development guide](docs/development.md) for repository orientation and the
-[release procedure](docs/release.md) for maintainer-only candidate and publication
-steps.
+## Design philosophy
+
+ConfigKit favors one coherent, predictable configuration model over flexibility
+for reproducing each consumer's conventions. Its goal is to let applications
+load once, inspect where final values came from, and distinguish explicit values
+from omissions with minimal configuration bookkeeping. Applications retain their
+domain rules and execution responsibilities.
+
+Loading is synchronous and complete on return. Applications normally load once
+near startup and then use the final configuration and metadata; ConfigKit stays
+inactive until explicitly called again. Applications may initiate additional
+loads as needed. See the [loading lifecycle](docs/config.md#completion-and-resource-lifetime)
+for completion and ownership details.
+
+Within a configuration tier, independent contributions compose strictly: shared
+mappings combine, while overlapping non-mapping definitions are errors. Between
+precedence tiers, mappings combine recursively, retaining omitted lower-priority
+children; other values replace lower-priority values. The main document and its
+imports form one strictly composed tier. Explicit underlay groups sit below it
+and above typed defaults; a selected overlay and explicit CLI overrides sit
+above it. The [configuration contract](docs/config.md#strict-document-composition)
+owns the detailed combination rules and input-specific constraints. At
+definition-registered paths only, strict composition can append records with
+distinct explicit IDs; ordinary and unregistered sequences keep their conflict
+or whole-value replacement behavior.
+
+The [architecture goals](docs/policy/architecture.md#success-criteria) explain
+these priorities and how genuine new needs can justify extending the shared
+model. The [future feature list](docs/roadmap/future.md) preserves candidates for
+consideration in future releases.
+
+The [configuration contract](docs/config.md) describes implemented behavior,
+including [input destinations](docs/config.md#input-destinations). The
+[consumer adoption guide](docs/consumers/README.md) provides maintained simple,
+layered, mounted-collection, and identified-prompt examples alongside the
+migration shapes. Input placement decisions are recorded in
+[ADR 0011](docs/adr/0011-input-destinations.md) and sequence placement and
+composition in [ADR 0014](docs/adr/0014-sequence-placement-and-composition.md).
+See the [development guide](docs/development.md) for repository orientation and
+the [release procedure](docs/release.md) for maintainer-only candidate and
+publication steps.
+
+ConfigKit code is licensed under the [LGPL v3.0 or later](COPYRIGHT).
+
+[Open Source attribution notices](docs/oss/attribution.md) are available.

conversion.go

@@ -43,16 +43,125 @@ func toEngineLocations(values []Location) []engine.Location {
 	return result
 }
 func fromEngineSourceInfo(value engine.SourceInfo) SourceInfo {
-	return SourceInfo{ID: value.ID, Label: value.Label, Path: value.Path, Layer: Layer(value.Layer), ImportedFrom: fromEngineLocationPointer(value.ImportedFrom)}
+	return SourceInfo{
+		ID: value.ID, Label: value.Label, Path: value.Path, Destination: fromEnginePath(value.Destination),
+		PlacementMode:         PlacementMode(engine.SourceInfoPlacementModeForFacade(value)),
+		DestinationUnresolved: engine.SourceInfoDestinationUnresolvedForFacade(value),
+		Layer:                 Layer(value.Layer), Kind: SourceKind(value.Kind), UnderlayIndex: value.UnderlayIndex,
+		ImportedFrom: fromEngineLocationPointer(value.ImportedFrom), VersionCheck: fromEngineSourceVersionCheck(value.VersionCheck),
+		ImportSelections: fromEngineImportSelections(engine.SourceInfoImportSelectionsForFacade(value)),
+	}
 }
 func toEngineSourceInfo(value SourceInfo) engine.SourceInfo {
-	return engine.SourceInfo{ID: value.ID, Label: value.Label, Path: value.Path, Layer: engine.Layer(value.Layer), ImportedFrom: toEngineLocationPointer(value.ImportedFrom)}
+	result := engine.SourceInfo{
+		ID: value.ID, Label: value.Label, Path: value.Path, Destination: toEnginePath(value.Destination),
+		Layer: engine.Layer(value.Layer), Kind: engine.SourceKind(value.Kind), UnderlayIndex: value.UnderlayIndex,
+		ImportedFrom: toEngineLocationPointer(value.ImportedFrom), VersionCheck: toEngineSourceVersionCheck(value.VersionCheck),
+	}
+	result = engine.SourceInfoWithImportSelections(result, toEngineImportSelections(value.ImportSelections))
+	return engine.SourceInfoWithPlacement(result, uint8(value.PlacementMode), value.DestinationUnresolved)
+}
+
+func fromEngineImportSelections(values []engine.ImportSelectionForFacade) []ImportSelection {
+	if values == nil {
+		return nil
+	}
+	result := make([]ImportSelection, len(values))
+	for index, value := range values {
+		result[index] = ImportSelection{
+			Path: fromEnginePath(value.Path), Location: fromEngineLocation(value.Location),
+			Selector: ImportSelector(value.Selector), Outcome: ImportSelectionOutcome(value.Outcome),
+		}
+	}
+	return result
+}
+
+func toEngineImportSelections(values []ImportSelection) []engine.ImportSelectionForFacade {
+	if values == nil {
+		return nil
+	}
+	result := make([]engine.ImportSelectionForFacade, len(values))
+	for index, value := range values {
+		result[index] = engine.ImportSelectionForFacade{
+			Path: toEnginePath(value.Path), Location: toEngineLocation(value.Location),
+			Selector: string(value.Selector), Outcome: string(value.Outcome),
+		}
+	}
+	return result
+}
+
+func toEngineImportDefaults(value ImportDefaults) engine.ImportDefaults {
+	return engine.ImportDefaults{
+		Required:  cloneImportDefaultBool(value.Required),
+		Follow:    cloneImportDefaultBool(value.Follow),
+		Hidden:    cloneImportDefaultBool(value.Hidden),
+		Recursive: cloneImportDefaultBool(value.Recursive),
+		Pattern:   cloneImportDefaultStrings(value.Pattern),
+	}
+}
+
+func cloneImportDefaultBool(value *bool) *bool {
+	if value == nil {
+		return nil
+	}
+	cloned := *value
+	return &cloned
+}
+
+func cloneImportDefaultStrings(values []string) []string {
+	if values == nil {
+		return nil
+	}
+	cloned := make([]string, len(values))
+	copy(cloned, values)
+	return cloned
+}
+
+func fromEngineSourceVersionCheck(value *engine.SourceVersionCheck) *SourceVersionCheck {
+	if value == nil {
+		return nil
+	}
+	return &SourceVersionCheck{
+		Path:     fromEnginePath(value.Path),
+		Version:  Version{Present: value.Version.Present, Null: value.Version.Null, Text: value.Version.Text},
+		Location: fromEngineLocationPointer(value.Location), Redacted: value.Redacted,
+	}
+}
+func toEngineSourceVersionCheck(value *SourceVersionCheck) *engine.SourceVersionCheck {
+	if value == nil {
+		return nil
+	}
+	return &engine.SourceVersionCheck{
+		Path:     toEnginePath(value.Path),
+		Version:  engine.Version{Present: value.Version.Present, Null: value.Version.Null, Text: value.Version.Text},
+		Location: toEngineLocationPointer(value.Location), Redacted: value.Redacted,
+	}
 }
 func fromEngineOrigin(value engine.Origin) Origin {
-	return Origin{Layer: Layer(value.Layer), Locations: fromEngineLocations(value.Locations), Operation: ChangeKind(value.Operation)}
+	result := Origin{Layer: Layer(value.Layer), Locations: fromEngineLocations(value.Locations), Operation: ChangeKind(value.Operation)}
+	if value.Inheritance != nil {
+		result.Inheritance = make([]InheritanceStep, len(value.Inheritance))
+		for index, step := range value.Inheritance {
+			result.Inheritance[index] = InheritanceStep{
+				From: fromEnginePath(step.From), To: fromEnginePath(step.To), Reference: fromEnginePath(step.Reference),
+				Locations: fromEngineLocations(step.Locations),
+			}
+		}
+	}
+	return result
 }
 func toEngineOrigin(value Origin) engine.Origin {
-	return engine.Origin{Layer: engine.Layer(value.Layer), Locations: toEngineLocations(value.Locations), Operation: engine.ChangeKind(value.Operation)}
+	result := engine.Origin{Layer: engine.Layer(value.Layer), Locations: toEngineLocations(value.Locations), Operation: engine.ChangeKind(value.Operation)}
+	if value.Inheritance != nil {
+		result.Inheritance = make([]engine.InheritanceStep, len(value.Inheritance))
+		for index, step := range value.Inheritance {
+			result.Inheritance[index] = engine.InheritanceStep{
+				From: toEnginePath(step.From), To: toEnginePath(step.To), Reference: toEnginePath(step.Reference),
+				Locations: toEngineLocations(step.Locations),
+			}
+		}
+	}
+	return result
 }
 func fromEngineDiagnostic(value engine.Diagnostic) Diagnostic {
 	return Diagnostic{Code: Code(value.Code), Severity: Severity(value.Severity), Path: fromEnginePath(value.Path), Message: value.Message, Primary: fromEngineLocationPointer(value.Primary), Related: fromEngineLocations(value.Related)}

development_check_test.go

@@ -9,8 +9,8 @@ import (
 	"testing"
 )
 
-func TestCheckScriptsHandleFilenamesAndStopOnFailure(t *testing.T) {
-	for _, scriptName := range []string{"release-check.sh", "check.sh"} {
+func TestDevelopmentCheckHandlesFilenamesAndStopsOnFailure(t *testing.T) {
+	for _, scriptName := range []string{"check.sh"} {
 		t.Run(scriptName, func(t *testing.T) {
 			script, err := os.ReadFile("scripts/" + scriptName)
 			if err != nil {
@@ -66,7 +66,7 @@ func TestCheckScriptsHandleFilenamesAndStopOnFailure(t *testing.T) {
 						t.Fatal(err)
 					}
 					cmd := exec.Command("/bin/bash", filepath.Join(root, "scripts", scriptName))
-					cmd.Env = append(os.Environ(), "PATH="+bin+string(os.PathListSeparator)+os.Getenv("PATH"), "CONFIGKIT_BASELINE_TOOLCHAIN=test-baseline", "CONFIGKIT_NEWER_TOOLCHAIN=test-newer", "CONFIGKIT_TEST_FMT_LOG="+log, "CONFIGKIT_TEST_GO_LOG="+calls)
+					cmd.Env = append(os.Environ(), "PATH="+bin+string(os.PathListSeparator)+os.Getenv("PATH"), "CONFIGKIT_TEST_FMT_LOG="+log, "CONFIGKIT_TEST_GO_LOG="+calls)
 					// Log only targets selected by the script, not inherited host settings.
 					cmd.Env = append(cmd.Env, "GOOS=", "GOARCH=")
 					if name == "failed-go-check" {
@@ -103,17 +103,6 @@ func TestCheckScriptsHandleFilenamesAndStopOnFailure(t *testing.T) {
 					if fail && strings.Contains(string(invoked), "vet ./...") {
 						t.Fatal("checker continued after failed test")
 					}
-					if scriptName == "release-check.sh" && !fail {
-						// These are selected release gates, not incidental loop details.
-						for _, target := range []string{"linux/amd64", "linux/arm64"} {
-							if !strings.Contains(string(invoked), target+" build ./...") {
-								t.Errorf("checker omitted required target %s", target)
-							}
-						}
-						if strings.Contains(string(invoked), "darwin/") {
-							t.Fatal("checker requires best-effort macOS validation")
-						}
-					}
 				})
 			}
 		})

diff.go

@@ -0,0 +1,53 @@
+package configkit
+
+import "gitea.maximumdirect.net/eric/ConfigKit/internal/engine"
+
+// Difference describes one canonical changed region between two successful
+// loads. Left is the earlier value and Right is the later value. Each snapshot
+// is detached; Present distinguishes absence from an explicit null, and
+// Redacted reports whether combined sensitivity affected that side. Mappings
+// are compared recursively, while sequences are reported as atomic values.
+// Records are ordered by canonical JSON pointer. HiddenChanged indicates that
+// a value concealed by combined sensitivity changed, without revealing which
+// concealed member changed.
+type Difference struct {
+	Path          Path
+	Left          ValueSnapshot
+	Right         ValueSnapshot
+	HiddenChanged bool
+}
+
+// Diff compares the frozen effective values of two successful loads of the
+// same configuration type. It does not reread sources, environment variables,
+// or rerun defaults or resolvers. Equal values produce no records even when
+// their sources, defaults, or resolution histories differ. Credential contents
+// are not compared: typed secret references compare by reference name only.
+//
+// Both results must be non-nil successful loads with inspection state; nil or
+// zero results return an error and no records. Load errors are returned by
+// Load, so callers must handle those before calling Diff. Snapshots are safe
+// for presentation but are not a complete reloadable configuration.
+func Diff[T any](left, right *Result[T]) ([]Difference, error) {
+	var leftInspection, rightInspection *engine.Inspection
+	if left != nil {
+		leftInspection = left.inspection
+	}
+	if right != nil {
+		rightInspection = right.inspection
+	}
+
+	values, err := engine.Diff(leftInspection, rightInspection)
+	if err != nil {
+		return nil, err
+	}
+	differences := make([]Difference, len(values))
+	for index, value := range values {
+		differences[index] = Difference{
+			Path:          fromEnginePath(value.Path),
+			Left:          fromEngineValueSnapshot(value.Left),
+			Right:         fromEngineValueSnapshot(value.Right),
+			HiddenChanged: value.HiddenChanged,
+		}
+	}
+	return differences, nil
+}

doc.go

@@ -2,7 +2,7 @@
 // loading for Go programs.
 //
 // Load supports file, byte, reader, and explicit no-input sources with strict YAML,
-// typed defaults, imports, profiles, CLI overrides, schema-version checks, detached
+// typed defaults, imports, overlays, CLI overrides, schema-version checks, detached
 // metadata, recorded application resolution, typed secret references, safe effective
 // rendering, read-only semantic validation, and structured load errors. The
 // configuration contract is documented in docs/config.md.

docs/adr/0005-explicit-overrides-and-safe-explanation.md

@@ -1,6 +1,8 @@
 # ADR 0005: Record explicit CLI overrides and safe value explanation
 
-Status: accepted
+Status: Partially superseded by
+[ADR 0018](0018-cli-after-structural-resolution.md) for override-before-decoding
+ordering. The remaining ownership and safe explanation decisions remain accepted.
 
 Date: 2026-09-06
 

docs/adr/0007-source-backed-underlays.md

@@ -0,0 +1,120 @@
+# ADR 0007: Add Source-Backed Underlays with Consistent Recursive Precedence
+
+## Status
+
+Proposed draft for review; its whole-sequence precedence replacement rule is
+superseded by [ADR 0017](0017-collection-sequence-policy.md). The remainder
+retains its proposed status. Completed planning and remediation records are
+preserved in Git history at `b604aa1` under `docs/roadmap/`; the
+[configuration contract](../config.md#layered-loading) owns current behavior.
+
+This extends [ADR 0001](0001-separate-composition-and-overlay.md) and preserves
+its distinction between strict composition and intentional overlay.
+
+## Date
+
+2026-09-29
+
+## Context
+
+Configuration-backed defaults provide lower-priority settings that applications
+can override. These sources carry imports, locations, and relative-resource
+contexts. Integrating them into ConfigKit's loading lifecycle gives their values
+the same presence tracking and inspectable provenance as other configuration
+inputs.
+
+The [architecture goals](../policy/architecture.md#success-criteria) prioritize
+one coherent, predictable configuration model with consistent rules across
+source representations and loading interfaces. Consumers should load once,
+receive a complete validated configuration, and inspect its final values' origins.
+Reducing integration code is important subject to those principles.
+
+## Decision
+
+### Add underlays to the existing precedence model
+
+Introduce explicitly supplied source-backed underlays between typed defaults
+and main configuration. The proposed precedence, from lowest to highest, is:
+
+**Typed application defaults < source underlays < strictly composed main
+configuration < selected overlay < explicit CLI overrides.**
+
+Typed defaults remain the Go-value baseline. Underlays supply lower-priority
+configuration with source provenance and, where applicable, imports and resource
+context. This ordering describes value precedence; typed defaults fill fields
+absent from the assembled input under the existing defaulting rules.
+
+Preserve strict composition of the main document and its imports before applying
+the resulting configuration over underlays. Independent contributions within an
+underlay tier also compose strictly. Explicitly ordered tiers express intentional
+precedence. Strict composition reports conflicting contributions within each tier.
+
+Across precedence tiers, reuse the existing recursive mapping-overlay semantics.
+Omitted children retain lower-priority input; existing scalar, sequence, null,
+and cross-kind replacement semantics remain unchanged. Reusing these operations
+keeps the combination rules consistent across the loading model.
+
+### Keep source representation separate from composition semantics
+
+Native `fs.FS` and programmatic inputs join the shared loading model.
+Equivalent inputs using the same
+operation have the same presence and precedence semantics regardless of medium.
+Programmatic values enter directly as labeled configuration sources.
+
+Retain the contributing source's identity and resource context through layering.
+A retained underlay value keeps its originating source's resource context.
+Consumers remain responsible for opening and interpreting their domain resources
+under the supplied filesystem's lifetime and access semantics. Exact
+context-access and input APIs remain roadmap decisions.
+
+### Deliver one final configuration and its inspection metadata
+
+Assemble effective input, then run the typed loading lifecycle once for the load.
+Preserve explicit zero,
+false, empty, and null input versus omission through configuration and metadata
+together.
+
+Extend existing provenance and history through underlays, defaults, and recorded
+application resolution. The returned result includes the completed combination,
+defaulting, and provenance work. Domain resolution and validation
+remain application-owned within the recorded lifecycle, consistent with
+[ADR 0003](0003-preserve-presence-and-resolution-history.md).
+
+Preserve the existing opt-in schema-version ownership, path protections, and
+check timing before precedence
+assembly and typed decoding. The roadmap owns how layered input identifies that
+owner when version checking is configured.
+
+Favor near-linear processing in input, retained history, and output size through
+incremental assembly of load-owned state. This is an engineering objective
+subordinate to clarity and semantic consistency. Preserve the root-owned facade and
+private implementation boundary established by [ADR 0006](0006-public-facade.md).
+
+## Alternatives considered
+
+- **Leave bundled defaults to consumers.** Avoids extending loading, but makes
+  consumers reproduce merging, presence tracking, and provenance reconciliation.
+  It does not meet the goal of obtaining a complete configuration in one load.
+- **Offer recursive or atomic combination per layer.** Makes both behaviors
+  explicit and could serve more consumers, but expands the semantic surface and
+  its interactions with presence and provenance without a demonstrated need that
+  outweighs a single consistent rule. A future concrete requirement may justify
+  reconsideration; this feature does not add the option.
+- **Flatten all documents into last-wins layers.** Simplifies ordering but turns
+  accidental main/import conflicts into overrides, weakening the intentional
+  distinction established by ADR 0001.
+
+## Consequences
+
+Consumers gain source-backed defaults within the shared loading lifecycle.
+Strict contributions and intentional precedence remain distinct, with consistent
+combination rules across source media. Presence and provenance remain
+part of the same loading lifecycle as the final values.
+
+Higher-priority mappings retain omitted lower-priority children, so application
+validation assesses the final combined configuration.
+
+Implementation must integrate source context and underlay history into the shared
+engine with attention to assembly and history-retention costs. Current behavior
+documentation will change when the feature lands. The roadmap retains ownership
+of the remaining API decisions.

docs/adr/0008-complete-on-return-loading.md

@@ -0,0 +1,82 @@
+# ADR 0008: Synchronous, Complete-on-Return Configuration Loading
+
+## Status
+
+Proposed draft for review. The
+[configuration contract](../config.md#completion-and-resource-lifetime) describes
+the implemented loading lifecycle and
+[source resource contexts](../config.md#source-resource-contexts).
+
+## Date
+
+2026-09-29
+
+## Context
+
+Applications need a clear point at which configuration is ready for use.
+Completing configuration processing and inspection metadata together gives
+applications a coherent result whose values and origins agree. A bounded loading
+lifecycle also makes resource ownership and startup failures straightforward to
+reason about.
+
+ConfigKit's [architecture goals](../policy/architecture.md#success-criteria)
+emphasize complete configuration, inspectable provenance, predictable behavior,
+and minimal consumer bookkeeping. Applications own their execution lifecycle and
+the timing of configuration acquisition.
+
+## Decision
+
+Make loading a synchronous, complete-on-return operation. Within each load,
+ConfigKit completes input assembly, defaulting, typed decoding, configured
+application resolution, secret lookup, validation, and snapshotting before
+returning the final configuration and metadata. Reported failures return
+diagnostics and a nil result.
+
+Release all library-owned loading handles before returning on success or failure.
+The resulting configuration and metadata are independently usable, caller-owned
+data with a retained inspection snapshot. Caller-supplied readers and filesystems
+remain under caller ownership and lifetime control.
+
+After the call completes, ConfigKit remains inactive until explicitly invoked
+again. Applications normally load once at or near startup and may initiate
+additional loads when needed. Each load follows the same lifecycle and source
+reuse contracts. Inspection, rendering, and separately requested operations run
+in response to explicit application calls.
+
+For the planned resource-context capability, retain source context as passive
+information or a borrowed caller-supplied filesystem. Subsequent resource access
+acquires its own handles with explicit ownership. This preserves a clear boundary
+between configuration completion and application resource use.
+
+These choices give consumers one completion boundary for configuration and
+provenance, predictable resource lifetimes, and control over when ConfigKit runs.
+Application callbacks remain responsible for their own effects and resources.
+
+## Alternatives considered
+
+- **Background watching or refresh.** Can automate responses to source changes,
+  but adds scheduling, shutdown, publication, and resource-lifetime concerns to
+  the library. Explicit calls keep those responsibilities with the application.
+- **Lazy completion after return.** Can defer work until particular values are
+  used, but moves configuration failures into application execution and makes
+  the readiness of values and metadata harder to establish.
+- **Transactional atomic loading.** Could imply simultaneous observation of all
+  external inputs and rollback of callback effects. Filesystems, environment
+  variables, and application callbacks do not share a transaction mechanism.
+  Synchronous completion provides the required usable-result boundary without
+  promising those external guarantees.
+- **Retain loading handles in results.** Can preserve open filesystem access,
+  but gives configuration results an ongoing resource lifecycle and cleanup
+  obligation. Passive source context supports later explicit resource access
+  while keeping loading resources scoped to the load call.
+
+## Consequences
+
+Applications can treat successful return as the point at which the final
+configuration and its metadata are ready for use. Startup loading remains simple,
+and additional loads have an explicit application-owned trigger.
+
+Configuration inspection operates on retained data. Resource access observes the
+backing resource when explicitly requested and has its own failure and ownership
+boundary. Implementations of new source kinds and layering features preserve the
+same completion boundary, including cleanup on reported load failures.

docs/adr/0009-sibling-inheritance.md

@@ -0,0 +1,82 @@
+# ADR 0009: Resolve Sibling Inheritance After Source Assembly
+
+## Status
+
+Partially superseded by [ADR 0017](0017-collection-sequence-policy.md) for
+sequence combination and the addition of identified-sequence sibling scopes,
+and [ADR 0018](0018-cli-after-structural-resolution.md) for CLI-before-inheritance
+ordering. Existing mapping-sibling scopes and the remaining ownership and
+inspection decisions remain accepted.
+
+The [configuration contract](../config.md#sibling-inheritance) owns implemented
+behavior. The completed implementation plan is preserved in Git
+history at `83c59ea`.
+
+## Date
+
+2026-09-29
+
+## Context
+
+Named configuration mappings often share settings while retaining individual
+overrides. ConfigKit already distinguishes strict composition of independent
+source contributions from recursive overlay between explicit precedence tiers.
+Doing another untracked merge in applications would lose omission, source
+context, and historical explanation information. A shared structural operation
+must preserve the same presence and provenance model as the rest of loading.
+
+Inheritance also interacts with source precedence, nested namespaces, version
+ownership, and secrets. Resolving it before source assembly would validate
+superseded references and copy values that a later tier should replace. Fully
+resolving nested relationships inside a reusable parent would bind them to the
+wrong sibling scope when copied into a child.
+
+## Decision
+
+Reserve `extends` as a built-in input key on mappings. A string names one exact
+sibling mapping; null cancels the relationship. Complete source acquisition,
+strict composition, selected overlay, and CLI precedence first. Then resolve
+each effective relationship in the parser-independent document tree, before
+typed defaults and decoding. A child recursively overlays its own payload on
+its resolved parent using the existing mapping-combination rules. This creates
+neither a source tier nor a configurable merge strategy.
+
+Resolve enclosing sibling chains before nested relationships. Reuse memoized
+parent templates with unresolved nested instructions; evaluate those instructions
+at each destination. Preserve original source applications separately from
+structural transfer/control evidence. Effective origins carry concrete
+source-to-destination derivation, and inspection projects relevant inheritance
+events between source replay and typed reconciliation. Inherited input counts as
+explicit presence, while direct replacement retains its own effective origin.
+
+The existing main-input version check remains at its original point. Reject
+structural contributions or directive consumption affecting its protected path.
+Classify sensitivity at corresponding source and destination paths and retain
+historical redaction. Secret references remain unresolved input until the
+existing lookup point. Keep all machinery behind the root-owned public facade.
+
+## Alternatives considered
+
+- **Application-owned inheritance after `Load`.** Leaves every consumer to
+  recreate ordering, presence, provenance, error reporting, and redaction after
+  ConfigKit has already declared a result complete.
+- **Resolve each source before precedence assembly.** Makes superseded
+  references active and can copy a parent definition before higher-priority
+  input finishes changing it.
+- **Resolve nested relationships in a cached parent subtree.** Shares more
+  ready-made output but binds nested sibling lookups to the parent's scope,
+  giving incorrect values at a child destination.
+- **Registered collections, configurable keywords, or multiple-parent merge
+  modes.** Increase public setup and semantic branches without a requirement
+  that warrants diverging from one structural rule.
+
+## Consequences
+
+Consumers can load a configuration once and inspect direct, inherited,
+defaulted, and application-derived values through the existing result surfaces.
+`extends` is no longer an ordinary application field, so existing configurations
+using that key as data must migrate. Expanding many children can materialize
+substantially more output than the input contains; projection and retained-history
+budgets bound that work without charging copies as newly acquired sources.
+Applications still own domain validation, entry selection, and runtime resource
+use.

docs/adr/0010-effective-configuration-diff.md

@@ -0,0 +1,99 @@
+# ADR 0010: Compare Frozen Effective Configuration Snapshots
+
+## Status
+
+Accepted and implemented. The
+[configuration contract](../config.md#effective-configuration-diffs) owns current
+behavior. Completed planning and remediation records are preserved in Git
+history at `1c2610c` under `docs/roadmap/`.
+
+## Date
+
+2026-10-09
+
+## Context
+
+Applications need to inspect how two completed configurations differ, including
+the effective impact of an overlay after defaults and application resolution.
+ConfigKit already retains canonical typed inspection snapshots, semantic paths,
+presence information, and sensitivity classification for successful loads.
+
+Extending this inspection model lets applications share configuration comparison
+and safe disclosure while keeping their presentation and domain interpretation.
+The operation should follow ConfigKit's existing equality, ownership, and secret
+boundaries so consumers can reason about loading and inspection consistently.
+
+## Decision
+
+Provide structured effective-value comparison between successful results of the
+same configuration type. Compare their retained canonical inspection snapshots
+through the root-owned public facade. Each comparison uses the completed load
+state and returns detached records with a semantic path and safe left and right
+value snapshots.
+
+Use canonical value equality and the existing structural comparison rules.
+Mappings compare recursively, sequences report atomic changes, and scalar or
+kind replacements report the affected region. Presence distinguishes additions
+and removals from explicit null. Equal effective values produce an empty diff;
+source attribution, explicit-input status, and history remain separately
+inspectable through each result. Canonical pointer ordering makes reporting
+deterministic.
+
+Detect changes before masking values. Apply sensitivity from both loads to both
+returned sides, including historical and dependency-derived classification.
+Preserve public content where the shapes permit corresponding sensitive regions
+to be identified. When a structural replacement prevents safe correspondence,
+mask the complete reported replacement region on both present sides. Absence
+retains its own representation.
+
+Keep masking and hidden-change information distinct. Each record indicates
+whether an actual change is withheld by redaction, allowing applications to
+display a changed-sensitive-value indicator even when safe snapshots render
+identically. Public-only changes alongside unchanged masked content retain that
+distinction.
+
+Typed secret references participate through their configured environment names.
+Resolved credential bytes remain behind the existing explicit secret accessor.
+The comparison uses the same frozen inspection boundary as effective rendering
+and explanation.
+
+ConfigKit owns comparison and paired safe disclosure. Applications own labels,
+layout, and rendering of the returned changed regions. Reusing the existing
+snapshot representation and engine boundary keeps the extension consistent with
+the library's established inspection model.
+
+## Alternatives considered
+
+- **Compare input text or separately rendered YAML.** Supports familiar textual
+  displays, but input text precedes composition and resolution, and independently
+  redacted output can hide changes or apply different disclosure rules. Canonical
+  comparison with paired masking represents the completed configuration.
+- **Leave comparison to applications using returned Go values.** Offers direct
+  access to application types, but duplicates structural and redaction machinery
+  and observes caller mutation. Retained inspection snapshots preserve the
+  completed load's values and established ownership boundary.
+- **Align sequence elements and return context-rich rendering data.** Enables
+  finer-grained displays, but requires additional correspondence and disclosure
+  rules. Atomic sequence records and existing value snapshots satisfy the
+  selected changed-region workflows with a smaller public model.
+- **Compare resolved credentials or retain credential-derived tokens.** Could
+  detect rotation behind an unchanged reference, but expands retained sensitive
+  state and introduces a separate trusted comparison mechanism. Reference-name
+  comparison preserves the existing inspection boundary.
+
+## Consequences
+
+Applications can compare independently loaded configurations using ConfigKit's
+shared equality, presence, and disclosure rules. Existing composition and load
+semantics continue to determine the values being compared. Caller mutation of a
+returned configuration leaves its retained comparison snapshot unchanged.
+
+Paired redaction can conceal public content within a structurally incompatible
+replacement. The changed region and hidden-change indicator still communicate
+its effective impact. Credential rotation behind an unchanged environment name
+produces equal reference values for comparison.
+
+The implementation requires operation-local disclosure preparation and focused
+coverage for presence, shape changes, historical sensitivity, and detached
+results. Applications retain control of presentation; changed-region records
+supply the safe data needed for that responsibility.

docs/adr/0011-input-destinations.md

@@ -0,0 +1,74 @@
+# ADR 0011: Place Configuration Inputs at Mapping Destinations
+
+## Status
+
+Accepted and implemented. Current placement behavior is owned by the
+[configuration contract](../config.md#input-destinations). The accepted
+relative-import extension is recorded in
+[ADR 0012](0012-relative-import-destinations.md). The completed roadmap, consumer
+request, and review records are preserved at `e4270c9` under `docs/roadmap/` in
+Git history; the original input-destination implementation plan remains at
+`315f464:docs/roadmap/implementation.md`.
+
+## Date
+
+2026-10-14
+
+## Context
+
+Consumers assembling collections of independently authored configuration
+fragments need each source to contribute beneath a selected application path.
+Requiring consumers to rewrite source documents or merge completed values would
+separate placement from ConfigKit's existing composition, precedence, source
+identity, and inspection behavior. Placement should remain part of the ordinary
+loading lifecycle and preserve the root-owned public facade.
+
+## Decision
+
+Add an optional `Destination Path` to root-owned `Input` and expose the effective
+destination on existing `SourceInfo` records. The empty path means the existing
+application root. A non-empty destination is a sequence of literal mapping keys;
+explicit sequence-index segments are invalid. Existing reserved structural keys
+remain reserved, while keys that merely contain their spelling remain ordinary
+keys.
+
+Acquire and interpret each source using its original physical identity and
+source-local directives, then place its application body before strict
+composition and the existing final lifecycle. Imported application bodies
+inherit the input destination. Placement applies to main and underlay inputs;
+ordinary root loading, selected overlays, and CLI inputs retain root placement.
+The main-owned version rule continues to check its effective application path
+at its existing lifecycle point.
+
+Placement preserves authored node locations and source identity. Introduced
+mapping structure belongs to the contributing source and uses that document's
+root location as its source-coordinate fallback. A present empty mapping
+materializes its destination; an absent `NoSource` contributes nothing. The
+result continues through the existing strict composition, overlay, inheritance,
+typed loading, diagnostics, provenance, history, sensitivity, and resource
+context mechanisms.
+
+The added exported fields preserve keyed and zero-value struct use. Callers
+using unkeyed composite literals for `Input` or `SourceInfo` must migrate to
+keyed literals. Treat this feature as minor-release work; select the version and
+document the compatibility impact during separately authorized release
+preparation, consistent with the release policy.
+
+## Alternatives considered
+
+- **Rewrite or merge sources in consumers.** This makes each consumer reproduce
+  placement, strict composition, source attribution, and inspection semantics.
+- **Add a separate loading entry point or public wrapper type.** This creates a
+  parallel API for a behavior that belongs to existing per-input loading.
+- **Load each fragment as a complete application before placement.** This runs
+  defaults, decoding, resolution, and validation on incomplete fragments rather
+  than once on the assembled application.
+
+## Consequences
+
+Consumers can assemble contributions at effective mapping paths in one ordinary
+layered load and inspect each source's placement through existing source
+records. Paths, conflicts, and later application operations use the placed
+structure while source coordinates and relative-resource contexts continue to
+identify their original sources. Public field additions require release-note
+guidance for callers using unkeyed literals.

docs/adr/0012-relative-import-destinations.md

@@ -0,0 +1,78 @@
+# ADR 0012: Place Imported Configuration at Relative Destinations
+
+## Status
+
+Accepted and implemented. The
+[configuration contract](../config.md#input-destinations) owns current behavior.
+The completed roadmap and review records are preserved at `e4270c9` under
+`docs/roadmap/` in Git history; the relative-import implementation plan remains
+at `d9f08ff:docs/roadmap/implementation.md`.
+
+## Date
+
+2026-10-14
+
+## Context
+
+ConfigKit can place an input document at a literal mapping path while retaining
+source identity, composition, and inspection in its normal loading lifecycle.
+An importing document can currently bring in other documents, but those
+contributions inherit only the importing input's destination. Consumers that
+assemble a collection therefore cannot declare the layout of imported
+fragments alongside the import itself without rewriting source documents or
+performing another merge outside ConfigKit.
+
+The extension should reuse input placement and the existing import lifecycle.
+Logical application placement must remain independent from physical path
+resolution, source identity, precedence, and resource ownership.
+
+## Decision
+
+Add an optional `destination` field to each YAML import declaration. Its value
+is a sequence of strings representing literal mapping-key segments. Omission
+and an empty sequence add no segments. A non-empty destination appends to the
+effective destination of the immediately importing document; nested imports
+continue from the resulting child destination. Every match of one glob
+declaration receives the same computed destination.
+
+Use the existing input-destination validation and reserved-key rules. Do not
+interpret destination segments as filesystem paths, separators, sequence
+indices, or root-reset instructions. Physical import paths continue to resolve
+from their declaring source under existing containment rules.
+
+Imports remain in their existing strict-composition group. A destination does
+not grant precedence, bypass cycle or repeated-source rules, or change source
+identity. Existing placement, version ownership, source records, diagnostics,
+resource contexts, inheritance, history, and inspection mechanisms continue to
+govern the placed contributions. No public Go entry point or independent
+specification-header behavior is introduced.
+
+This is additive YAML functionality and is minor-release work. Version
+selection, release notes, roadmap retirement, and publication remain separate
+release-preparation work.
+
+## Alternatives Considered
+
+- **Require complete application-root destinations on every import.** This
+  avoids inherited context but repeats collection prefixes in nested source
+  declarations and does not express placement relative to the importing
+  document.
+- **Encode destinations as dotted or slash-delimited strings.** This introduces
+  separator and escaping rules even though existing paths already represent
+  literal mapping keys as segments.
+- **Rewrite or merge imported documents in consumers.** This duplicates source
+  placement, strict composition, attribution, and inspection behavior outside
+  ConfigKit's loading lifecycle.
+- **Add independent specification headers or collection-specific validation.**
+  These are separate design questions and are not required to declare relative
+  placement.
+
+## Consequences
+
+Import declarations can describe the logical placement of their contributions
+while preserving the physical source and existing composition lifecycle.
+Placement is immediate-parent-relative, so nested imports accumulate their
+declared segments and sibling declarations remain independent. The feature is
+implemented behind the existing public loading surface; detailed validation,
+diagnostic, accounting, and inspection requirements are tracked by the
+implementation roadmap.

docs/adr/0013-shared-source-version-checks.md

@@ -0,0 +1,92 @@
+# ADR 0013: Check Source Versions with One Shared Policy
+
+## Status
+
+Accepted and implemented. The
+[configuration contract](../config.md#shared-source-version-checks) owns current
+behavior. The completed roadmap, implementation plan, consumer request, and
+resolved review findings are preserved at `e4270c9` under `docs/roadmap/` in
+Git history; the [development guide](../development.md#baseline-validation)
+provides retrieval examples.
+
+## Date
+
+2026-10-15
+
+## Context
+
+Applications that assemble one configuration from independently authored
+documents need a way to require that each contributing document declares a
+version accepted by the application's existing `Definition.Version` checker.
+Checking only the eventual composed value lets another source hide a rejected
+declaration, while separate consumer-side parsing duplicates ConfigKit's source
+handling and diagnostics.
+
+The extension must preserve one application-owned acceptance policy and the
+existing loading lifecycle. Source declarations are compatibility evidence, not
+independent fragment schemas or values to compare for equality. Existing
+composition, placement, precedence, source identity, diagnostics, and
+sensitivity rules remain authoritative.
+
+## Decision
+
+Provide a load-wide source-version mode for ordinary and layered loading. The
+zero-value mode preserves current behavior. The enabled modes check declarations
+when present or require them, using the existing `Definition.Version` path and
+checker. Interpret the configured path relative to each participating source
+before placement or composition. Different declarations may be accepted by the
+same checker; ConfigKit adds no equality, SemVer, or expected-version rule.
+
+Apply the selected policy to the actual main document, direct underlays,
+imports, and selected overlays. Missing non-main declarations are skipped in
+if-present mode and rejected in required mode. Existing main absence behavior
+remains, with required mode additionally rejecting an absent declaration on an
+actual main document. `NoSource`, unmatched optional inputs, synthetic field
+operations, and empty groups do not become documents that need declarations.
+
+Reuse the existing version vocabulary and validation boundary: absence, null,
+exact strings, and strict decimal integers retain their established meanings;
+unsupported present values fail version-type validation. Warnings remain
+nonfatal, and source declarations do not replace final application decoding or
+domain validation.
+
+After an accepted non-main declaration is checked, consume exactly that node
+before its remaining application content is composed or overlaid. Preserve the
+main declaration and its value semantics. Keep the declaration, source-local
+path, physical location, and checked absence as detached source evidence,
+separate from effective application values and history. Apply established
+sensitivity rules to that evidence, including original mounted paths and
+ancestors; retain existing safe failure disclosure.
+
+In enabled modes, the main declaration remains source-local while its protected
+application path is derived from the main input's destination. This distinction
+does not change default-mode path interpretation or weaken protection of
+remaining payload and synthetic operations.
+
+Public option-field additions and the enabled-mode interpretation for mounted
+main inputs are compatibility considerations for separately authorized minor-
+release preparation. This decision does not select a release version or
+authorize release notes or publication.
+
+## Alternatives Considered
+
+- **Check only the final composed value.** Later sources can hide an earlier
+  rejected declaration, so this does not validate every contributing document.
+- **Let consumers parse and check each source.** This duplicates acquisition,
+  source attribution, safe diagnostics, and lifecycle coordination in every
+  application.
+- **Require declarations to be equal or impose a universal version grammar.**
+  Consumers already own the acceptance policy; ConfigKit need not invent an
+  expected value or a domain-specific compatibility rule.
+- **Keep accepted non-main declarations in application payload.** A selected
+  source-local declaration is metadata for this policy and cannot also remain
+  an ordinary field at the same path without ambiguity.
+
+## Consequences
+
+Consumers can apply one existing version policy across contributing documents
+and select gradual or required adoption per load. Accepted non-main headers no
+longer participate in application composition; their inspectable evidence is
+kept separate and subject to sensitivity classification. Default callers retain
+current behavior. The implementation plan records the concrete API names,
+integration order, and validation evidence.

docs/adr/0014-sequence-placement-and-composition.md

@@ -0,0 +1,81 @@
+# ADR 0014: Place and Compose Sequence Items Explicitly
+
+## Status
+
+Partially superseded by [ADR 0017](0017-collection-sequence-policy.md) for
+identity ownership and sequence combination. Explicit source placement remains
+accepted.
+
+The [configuration contract](../config.md#strict-document-composition)
+defines implemented behavior, and the [consumer guide](../consumers/README.md#identified-prompt-sequences)
+shows the maintained example. The completed roadmap, implementation plan, and
+resolved remediation record are preserved in Git history at `008d2ba` under
+`docs/roadmap/`; see the [development guide](../development.md) for retrieval
+commands.
+
+## Date
+
+2026-10-16
+
+## Context
+
+ConfigKit strictly composes independent documents by recursively combining
+mappings and rejecting other overlapping values. This makes two independently
+authored sequence fragments conflict, even when their records are distinct.
+Precedence assembly has a different purpose: a higher-priority present sequence
+replaces the lower-priority sequence as one value. Those existing rules are
+predictable, but they do not express the separate operation of collecting
+independent records while rejecting duplicate identities.
+
+Likewise, mapping destinations place a source beneath literal mapping keys, but
+there is no explicit way to state that a source mapping represents one item in a
+sequence. Inferring that intent from document shape or fields would make source
+meaning implicit and change existing placement behavior.
+
+## Decision
+
+Keep mapping placement as the zero-value behavior and add an explicit item
+placement choice for a source whose mapping should become one element at a
+non-root sequence destination. Item placement provides structure only; it does
+not infer or synthesize identity values.
+
+Allow a definition to register literal mapping paths and one literal identity
+field for sequences at those paths. Within each strict composition tier, a
+registered sequence accepts mapping records with a present string identity,
+appends distinct records in contribution order, and rejects repeated exact
+identity strings. Identical registrations coalesce; conflicting identity fields
+for the same path are invalid. Registration does not recursively select paths
+inside sequence records.
+
+Keep precedence assembly unchanged. A present higher-priority sequence replaces
+the lower-priority value wholesale, including at a registered path. Registration
+changes strict composition only; it does not turn precedence into append or
+identity-based merging.
+
+This decision preserves ConfigKit's distinction between independent
+same-priority contributions and intentional precedence. Explicit source
+placement and definition-owned composition policy make the two required choices
+visible while keeping the existing zero-value load behavior.
+
+## Alternatives Considered
+
+- **Infer item placement from a source mapping or identity field.** Shape-based
+  inference would make ordinary mapping placement depend on document contents
+  and would leave the intended destination ambiguous.
+- **Append all overlapping sequences during strict composition.** This would
+  change existing conflict semantics at unregistered paths and would not detect
+  duplicate records.
+- **Merge matching records recursively by identity.** This would add a second
+  recursive merge rule with different conflict and precedence behavior from
+  ConfigKit's existing mapping composition and overlay operations.
+- **Use identity to merge sequences across precedence tiers.** This would
+  change established whole-value replacement semantics and conflate strict
+  composition with precedence.
+
+## Consequences
+
+Consumers can explicitly place mappings as sequence items and opt selected
+sequences into duplicate-rejecting strict composition. Unregistered sequences
+retain their current strict-conflict behavior, and higher-priority sequences
+continue to replace lower-priority ones. The implementation plan owns the exact
+API, YAML syntax, diagnostics, and implementation status.

docs/adr/0015-shared-dependency-ready-ordering.md

@@ -0,0 +1,68 @@
+# ADR 0015: Share Dependency-Ready Ordering Across Imports and Inheritance
+
+## Status
+
+Accepted. Implemented behavior is described in the
+[configuration contract](../config.md#explicit-imports) and
+[internal import overview](../internal/imports.md).
+
+## Date
+
+2026-10-19
+
+## Context
+
+Imports and sibling inheritance both resolve relationships between configuration
+contributions. Their existing resolvers use different traversal rules, making
+ordering depend on which mechanism expresses a dependency. Ordering affects
+sequence contributions, provenance, and diagnostics, so a consistent rule is
+part of ConfigKit's configuration model.
+
+The resolvers also serve distinct purposes. Imports strictly compose sources;
+inheritance overlays a child on its resolved parent after source assembly.
+Item-local contributions and nested inheritance introduce scope boundaries that
+resolution must preserve.
+
+## Decision
+
+ConfigKit owns one shared dependency-ready scheduler for imports and sibling
+inheritance. Both resolvers use Kahn's algorithm with a stable ordering key,
+selecting the smallest currently eligible node after its dependencies have
+completed. The shared scheduler owns eligibility, deterministic tie-breaking,
+and dependency-cycle identification. Each resolver owns construction of its
+dependency relationships and application of its domain operations.
+
+Apply the common scheduling rule within the scopes where dependencies are
+available. Complete item-local contributions before item placement and resolve
+enclosing inheritance before relationships in the resulting nested scopes.
+Source discovery supplies the import graph; the scheduler determines resolution
+order.
+
+Strict composition, inheritance overlay, and precedence tiers retain their
+existing ownership and meanings under [ADR 0001](0001-separate-composition-and-overlay.md).
+The lifecycle and scope decisions in
+[ADR 0009](0009-sibling-inheritance.md) continue to govern inheritance. Import
+admission retains its cycle and repeated-source rules.
+
+## Alternatives Considered
+
+- **Deterministic depth-first postorder.** Keeps dependency branches together,
+  but gives traversal branch selection an additional role in ordering.
+- **Dependency rounds.** Groups nodes by readiness rounds, but delays newly
+  eligible nodes until the current round finishes.
+- **Descending longest-path depth.** Prioritizes the furthest dependencies,
+  but adds distance from the root as an ordering policy.
+- **Separate resolver algorithms.** Allows local specialization, but leaves
+  consumers and maintainers with different rules for analogous dependencies.
+
+## Consequences
+
+Both resolvers follow one deterministic ordering model and share responsibility
+for maintaining that model. Resolver-specific scopes, combination semantics,
+and source-admission rules remain explicit at their integration boundaries.
+
+Independent branches may interleave as nodes become eligible. Adopting this
+rule changes existing traversal-dependent sequence positions and reporting
+order, requiring coordinated review of source assembly, provenance, and
+inheritance behavior. The configuration contract owns the implemented ordering
+details.

docs/adr/0016-independent-import-controls.md

@@ -0,0 +1,83 @@
+# ADR 0016: Separate Import Selection, Traversal, and Presence Requirements
+
+## Status
+
+Accepted. Implemented behavior is described in the
+[configuration contract](../config.md#explicit-imports), the
+[public API declarations](../../lifecycle.go), and the
+[internal import overview](../internal/imports.md).
+
+## Date
+
+2026-10-19
+
+## Context
+
+Consumers assemble configurations from named files and collections whose layout
+varies between installations. Simple imports should require little setup, while
+larger collections need control over discovery, recursion, filtering, and
+required contributions. Combining those choices implicitly makes declarations
+harder to interpret and pushes source-selection work into applications.
+
+Discovery can also reach a candidate through overlapping searches or encounter
+different names for one canonical source. Selection and source admission need
+distinct identities to support useful search unions while preserving ConfigKit's
+existing source-integrity rules.
+
+## Decision
+
+ConfigKit owns import discovery and selection within its source-loading
+lifecycle. Support explicit files, directory searches, and path globs as
+distinct selection forms. Keep traversal, filename filtering, and presence
+requirements independently expressible. Discovery determines the selected
+contributions; the shared scheduler in
+[ADR 0015](0015-shared-dependency-ready-ordering.md) determines their resolution
+order.
+
+Use optional-by-default imports with explicit requiredness controls. Selected
+sources remain subject to ConfigKit's validity, containment, resource, and
+composition contracts. This applies the library's
+[simplicity and defaults principles](../policy/architecture.md#simplicity-defaults-and-explicit-controls)
+to import absence while retaining meaningful errors for invalid contributions.
+
+Applications own their selected import defaults. Authored declaration values
+take precedence over application defaults, which take precedence over library
+defaults. Each declaration resolves applicable controls independently,
+including declarations in imported sources. This gives applications a shared
+baseline while preserving explicit choices in individual declarations.
+
+Keep declaration-local candidate identity separate from canonical source
+identity. Search unions coalesce repeated candidate paths within a source
+namespace before symlink resolution; filename filtering uses the selected or
+discovered name. Source admission continues to use canonical identity and the
+existing repeated-source rules. Distinct aliases retain their separate
+inclusion routes.
+
+## Alternatives Considered
+
+- **Use path globs for all discovery.** Offers one selection form, but makes
+  collection traversal depend on path-pattern structure and complicates
+  independently controlling recursion and filename filtering.
+- **Offer only explicit files and directory searches.** Reduces the selector
+  vocabulary, but moves wildcard directory selection into repeated declarations
+  or application-side discovery.
+- **Require every import by default.** Detects absent contributions immediately,
+  but requires repeated opt-outs for installations with optional components.
+- **Make application settings override authored declarations.** Provides central
+  enforcement, but prevents declarations from expressing local exceptions.
+- **Deduplicate selections globally by canonical source identity.** Coalesces
+  aliases and separate inclusion routes, but changes the existing repeated-source
+  contract and obscures independently authored contributions.
+
+## Consequences
+
+Consumers can express simple imports and more selective collection searches
+through the same loading lifecycle. Applications choose defaults and validate
+their required effective configuration; ConfigKit owns consistent selection,
+source admission, composition, and inspection.
+
+Optional absence can conceal misspelled selectors, so requiredness and
+inspection provide explicit ways to detect and explain missing contributions.
+Declaration-local deduplication requires the loader to preserve candidate
+identity alongside canonical source identity. The configuration and public API
+contracts own the implemented behavior.

docs/adr/0017-collection-sequence-policy.md

@@ -0,0 +1,143 @@
+# ADR 0017: Define Identity and Merge Rules for Each Collection
+
+## Status
+
+Accepted.
+
+This supersedes the decisions about who declares identity and how
+sequences combine in [ADR 0014](0014-sequence-placement-and-composition.md),
+and the rule that inheritance always replaces sequences in
+[ADR 0009](0009-sibling-inheritance.md). It also extends that ADR's sibling
+scopes to identified sequences. Existing mapping-sibling scopes and source
+placement continue to apply. It supersedes the whole-sequence
+precedence replacement rule in the proposed
+[ADR 0007](0007-source-backed-underlays.md).
+
+## Date
+
+2026-10-22
+
+## Context
+
+ConfigKit can load separate files as records in a sequence. Applications can
+register an identity key so records with different identities combine and
+duplicates fail. However, a higher configuration layer or an inheriting mapping
+currently replaces the whole sequence. Applications that need to extend a
+catalog across layers must write their own merging code and keep source and
+history information consistent with its results.
+
+Configuration authors need to choose how records are identified and whether a
+layer adds records, updates records, or replaces the collection. Applications
+also need a way to require unique identities in the final loaded configuration.
+These rules must work consistently for imported records, inline sequences,
+overlays, and inheritance.
+
+Records in a sequence may also share settings. Authors need to reuse those
+settings through inheritance while keeping each record's own identity.
+
+## Decision
+
+ConfigKit manages identity and merge rules for each collection during loading.
+Configuration files can declare those rules for imported and inline sequences,
+including sequences in selected overlays. Applications can require a particular
+identity definition, and configuration declarations must agree with it. Source
+placement determines where each contribution enters the configuration.
+
+Keep the distinction between contributions at the same priority and overrides
+from a higher layer, as established in
+[ADR 0001](0001-separate-composition-and-overlay.md). Within one tier, records
+with distinct identities combine; duplicate identities and conflicting
+declarations fail. Determine the applicable rules before combining those
+contributions, including any identity established in lower tiers.
+
+Across layers, support three operations: replace the whole sequence, append
+records while rejecting duplicate identities, or replace whole records by
+identity (`upsert`). Source layering and sibling inheritance use the same
+operations. For inheritance, the rules at the inheriting destination determine
+which operation applies.
+
+An identified sequence is also a sibling scope for inheritance. A record may
+inherit from one other record in the same sequence, selected by its complete
+identity. Each record supplies its own identity before inheritance. The child
+keeps that identity and combines its settings with the parent's other fields,
+using the existing mapping and collection rules. Parents remain ordinary records
+subject to normal validation.
+
+Resolve these relationships after source assembly, using the same dependency
+ordering as mapping-sibling inheritance. Resolve nested relationships at each
+inheriting destination. Preserve the assembled sequence order. Whole-record
+upsert keeps its existing meaning; inheritance then uses the records that remain.
+Typed defaults, CLI overrides, and application resolution follow the lifecycle
+in [ADR 0018](0018-cli-after-structural-resolution.md).
+
+An explicitly selected operation remains in effect at its destination until a
+higher layer changes it. When no explicit operation applies, use upsert if an
+identity definition is available; otherwise replace the whole sequence. Authors
+can declare identity and an operation before supplying any records. Keep source
+acquisition order separate from layer priority and the dependency ordering
+established by [ADR 0015](0015-shared-dependency-ready-ordering.md).
+
+Rules declared inside a record belong to that record. Retaining the record
+retains its local rules; replacing it discards them. Replacing a whole collection
+can discard its previous YAML identity definition, while application requirements
+still apply. Inheritance can copy collection identity, but the inheriting
+destination chooses its operation. Application registrations continue to govern
+the paths where they were registered. The same ownership rules apply to
+declarations made before a sequence exists.
+
+Check YAML-declared identity when combining collections during source assembly
+and inheritance. Application registrations also require the final loaded
+sequences to satisfy their identity rules, including after typed defaults, CLI
+overrides, and application resolution. This guarantee checks sequences that
+exist; the application schema and validation decide whether a collection must
+exist and which value types are allowed. The guarantee applies when loading
+completes, rather than to every intermediate callback state or changes made by
+the caller afterward.
+
+Preserve explicit presence, source attribution, history, sensitivity, and
+resource context through these operations, following
+[ADR 0003](0003-preserve-presence-and-resolution-history.md). Typed defaults,
+CLI assignments, and application resolution keep their own rules for changing
+values. The public API provides registration and inspection; the implementation
+chooses how to store and track collection rules internally.
+
+## Alternatives Considered
+
+- **Keep identity application-owned and always replace sequences across layers.**
+  Keeps the current design small, but reusable catalogs require application
+  changes or custom merging code.
+- **Put merge rules on each import.** Keeps the rules beside the imported source,
+  but inline sequences and overlays need a separate way to declare them. It also
+  makes rules for a collection spread across several imports harder to reconcile.
+- **Merge individual fields of records with matching identities.** Allows partial
+  record updates, but adds rules for field conflicts and retained values.
+  Replacing the whole record makes the update's scope clear.
+- **Keep inheritance limited to mapping siblings.** Keeps parent lookup simpler,
+  but authors of sequence records must duplicate shared settings or applications
+  must implement inheritance themselves. Collection identity provides a way to
+  select a parent within the existing loading process.
+- **Check every YAML identity again at the end of loading.** Gives YAML declarations
+  authority over values changed by defaults, CLI assignments, and application
+  resolution. This extends the lifetime of those declarations; application
+  registration provides that final guarantee when needed.
+
+## Consequences
+
+Consumers can build and customize catalogs in one load while retaining useful
+inspection information. Configuration authors choose how collections combine
+within application requirements. Duplicate identities among contributions at
+the same priority remain errors.
+
+Existing registered collections use upsert across layers by default unless an
+explicit operation selects otherwise. Correct loading now requires tracking
+which declarations still apply as records are replaced, copied, or moved to new
+positions. Lower-tier identity must be available before combining higher-tier
+contributions.
+
+Record inheritance also requires parent lookup by identity and checking for
+missing parents and cycles. Copied settings retain the same source, history,
+and sensitivity information as mapping-sibling inheritance.
+
+The configuration contract will define declaration syntax, identity comparison,
+omission and empty-value behavior, and diagnostics when the feature is
+implemented.

docs/adr/0018-cli-after-structural-resolution.md

@@ -0,0 +1,84 @@
+# ADR 0018: Apply CLI Overrides After Inheritance and Decoding
+
+## Status
+
+Accepted.
+
+This supersedes the rule that CLI overrides apply before decoding in
+[ADR 0005](0005-explicit-overrides-and-safe-explanation.md) and the
+rule that they apply before inheritance in [ADR 0009](0009-sibling-inheritance.md).
+Their remaining ownership, inheritance scope, and safe inspection decisions
+continue to apply.
+
+## Date
+
+2026-10-22
+
+## Context
+
+Explicit command-line values are the operator's highest-priority input. If they
+apply before inheritance, a sequence supplied as a replacement could gain
+inherited records afterward. Its final contents would then differ from the
+operator's assignment.
+
+The application also needs these values when it fills missing settings,
+normalizes values, or computes derived settings. ConfigKit needs a clear loading
+order that preserves both CLI precedence and the application's role in resolving
+the final configuration.
+
+## Decision
+
+ConfigKit uses this loading order:
+
+1. Combine sources within tiers and apply precedence between layers.
+2. Resolve sibling inheritance, apply typed defaults, and decode the configuration
+   into the application's types.
+3. Apply explicit CLI overrides and check them against the application schema.
+4. Run application resolution using the overridden configuration.
+5. Look up effective secret references, validate the final configuration, and
+   return it.
+
+CLI overrides are the highest-priority input to application resolution. A CLI
+sequence replaces the whole sequence; mappings combine recursively. Collection
+merge rules apply during the earlier source and inheritance steps. Inheritance
+and typed defaulting do not run again after CLI overrides.
+
+Keep track of which values were explicitly supplied and which CLI input supplied
+them. Applications continue to parse command-line arguments; ConfigKit applies
+and validates the overrides and makes their effects inspectable. Reject
+structural directives in override paths and values, so instructions such as
+inheritance relationships remain part of source processing.
+
+Application resolution keeps its existing rules for defaults and recorded
+changes under [ADR 0003](0003-preserve-presence-and-resolution-history.md).
+It receives CLI values together with their presence and source information.
+Defaults that only fill missing settings preserve explicit CLI values.
+Application normalization and derived-value calculations can still change them,
+with those changes recorded. Secret lookup and final validation use the resulting
+configuration.
+
+## Alternatives Considered
+
+- **Apply CLI overrides before inheritance.** Preserves existing ordering, but
+  later merging needs special handling to preserve CLI replacement intent while
+  still combining ordinary source contributions.
+- **Apply CLI overrides after inheritance but before typed decoding.** Allows an
+  override to fix a source value that would otherwise fail decoding, but leaves
+  typed defaulting after the operator's assignment. Completing defaulting and
+  decoding first gives CLI overrides an already resolved configuration to modify.
+- **Apply CLI overrides after application resolution.** Applies assignments after
+  the application's changes, but derived settings still reflect earlier values
+  unless resolution runs again.
+
+## Consequences
+
+CLI overrides change a configuration whose inheritance and typed defaults are
+already resolved. The application can then derive settings from the operator's
+values. The final values, source information, history, and sensitivity handling
+must remain consistent throughout loading.
+
+CLI input can no longer add or cancel inheritance relationships or bypass an
+earlier error in source assembly or inheritance. It also cannot fix a value that
+fails the earlier decoding step. Semantic validation still runs after overrides
+and application resolution. The configuration contract and consumer guidance
+must explain these behavior changes when they are implemented.

docs/config.md

@@ -12,20 +12,30 @@ ConfigKit exposes opaque sources through these constructors:
 | Constructor | Current contract |
 | --- | --- |
 | `File(path)` | Requires a nonempty path, opens a regular file for each consumption, and closes only the library-owned file handle. Missing, inaccessible, unreadable, directory, and other nonregular inputs fail. |
+| `NewFilesystem(label, filesystem).File(name)` | Borrows a caller-owned `fs.FS` under a safe nonempty display label. `name` must be a non-root slash-relative `fs.ValidPath`. Each load reopens the selected regular file; copies of the `Filesystem` preserve namespace identity. Separate constructors create separate namespaces even for the same label or filesystem. |
 | `Bytes(label, data)` | Requires a nonempty, non-secret diagnostic label and snapshots bytes during construction. Later caller mutation cannot affect the source. |
 | `Reader(label, input)` | Requires a nonempty, non-secret label and non-nil reader. The reader is claimed once across all copies of the source, consumed but never closed, and cannot be reused. |
+| `Mapping(label, values)` | Requires a safe nonempty label and a non-nil root `map[string]any`. Snapshots supported built-in values into an owned mapping document at construction; `{}` is valid. Constructor conversion failures become safe deferred diagnostics. |
 | `NoSource()` | Selects defaults-only loading, creates no source record, and marks no paths explicit. |
 
 The zero `Source` is invalid. Constructor argument failures are retained inside the
 opaque source and become safe structured diagnostics when the source is consumed.
-File sources can observe file changes between consumptions; byte sources cannot.
+File and filesystem sources can observe changes between consumptions; byte and
+mapping sources cannot. Mapping accepts nil, built-in string, boolean, signed or
+unsigned integer, finite float, `[]any`, and `map[string]any` values recursively.
+Typed nil maps and slices represent explicit null; empty collections remain empty.
+Keys are sorted for stable provenance, integer decimal spelling stays exact, and
+cycles or unsupported types are rejected without exposing values. Mapping input
+uses the same directive and version rules as YAML, not CLI override restrictions.
 
-Each source is limited to 8 MiB. A load using imports or an active profile is limited
+Each source is limited to 8 MiB. A load using imports or an active overlay is limited
 to 64 MiB across all accepted sources, 1,024 accepted sources, 64 active import
 edges, and 1,000,000 input nodes in aggregate. Glob expansion separately permits
 1,000,000 visited directory entries, including nonmatches. Reads request at most the
 remaining allowance plus one byte so excess data is rejected rather than silently
 truncated. Data returned together with a read error is not accepted as partial input.
+For mapping sources, key and string bytes use the same accounting as programmatic
+overrides, and the converted tree uses the same node and depth bounds.
 
 ## Single-source loading
 
@@ -33,7 +43,7 @@ truncated. Data returned together with a read error is not accepted as partial i
 root `T` must satisfy the application schema contract below. With no options it
 starts from the Go zero value. A `Definition[T].Defaults` factory, when supplied,
 runs exactly once after successful source parsing, schema construction, optional
-schema-version checking, and selected-profile loading. Its result is detached before
+schema-version checking, and selected-overlay loading. Its result is detached before
 input is applied. Resolver-provided mutable values are detached before secret lookup
 and validation, and a final detached copy prevents equal-valued aliases introduced
 by validators from escaping into the returned configuration. The publication copy
@@ -43,34 +53,253 @@ The current lifecycle is:
 
 ```text
 validate options → parse the selected source → traverse/compose enabled imports
-  → build schema → check the optional application version → apply a selected profile
-  → apply explicit CLI overrides → create and clone defaults → strictly decode
+  → build schema → check the optional application version → apply a selected overlay
+  → apply explicit CLI overrides → expand sibling inheritance
+  → create and clone defaults → strictly decode
   → recorded application resolution
   → resolve typed secret references → semantic validation → classify secrets
   → snapshot configuration and metadata
 ```
 
-`File`, `Bytes`, and `Reader` each contribute the first base source with deterministic
+`File`, `Filesystem.File`, `Bytes`, `Reader`, and `Mapping` each contribute the first base source with deterministic
 ID `source-1`; enabled imports continue in depth-first order as `source-2` onward.
 `NoSource` skips parsing, supplies an absent version to a configured version checker,
 contributes no source, and continues through defaults, decode, and projection. A
 missing file remains an error; it is never interpreted as no input.
 
-The current `Options` boundary accepts a defaults factory, `VersionRule`, resolver,
-semantic validator, explicit path and glob imports, one selected profile, and
-filesystem context. It also accepts explicit CLI override records, static secret
-tags, and a dynamic path predicate. `NoSource` rejects imports, active profiles,
-and filesystem context as invalid combinations but accepts overrides. Input
-containing a reserved `configkit` mapping key is rejected while imports are
-disabled.
+`Options` accepts the application definition, ordinary-source `Destination` and
+`PlacementMode`, explicit imports, one selected overlay, filesystem
+context, explicit CLI overrides, and shared source-version controls. The definition
+may register strict sequence composition paths through
+`SequenceCompositions`. Its zero-valued placement mode places the source mapping
+at the destination; `sequence-item` explicitly wraps the mapping as one item at a
+non-root destination. Unknown modes, invalid destinations, and invalid or
+conflicting sequence registrations fail before source acquisition. Overlay and
+CLI coordinates remain rooted at the application root. `NoSource` rejects
+imports, active overlays, and filesystem context as invalid combinations but
+accepts overrides. Input containing a reserved `configkit` mapping key is rejected
+while imports are disabled.
 
 Every source, options, schema, version, defaults, decode, resolution, or validation
 failure returns a nil result and a `*LoadError`. Successful results contain caller-owned `Config`, detached
-`Metadata`, a defensive source list, and defensive nonfatal diagnostics. Mutating
+`Metadata`, a defensive source list, and defensive nonfatal diagnostics. Each source
+record's `Kind` identifies file, filesystem, bytes, reader, mapping, or CLI acquisition
+independently of its precedence `Layer`. Host file paths remain canonical;
+filesystem paths are slash-relative within their namespace; synthetic inputs have
+no path. Mutating
 the returned configuration or inspection records cannot alter ConfigKit's private
 snapshot. ConfigKit performs no logging. `Result.WriteEffective` exposes only the
 frozen redacted projection.
 
+### Completion and resource lifetime
+
+`Load` is synchronous and complete on return. Success means input assembly,
+defaulting, typed decoding, configured application resolution, secret lookup,
+validation, and configuration/metadata snapshotting have completed. A reported
+load failure returns diagnostics and a nil result, as described above.
+
+Before returning on success or failure, ConfigKit closes the filesystem handles
+it opened for loading, including files and directory/root handles. Caller-owned
+filesystems are borrowed, never closed. The returned
+configuration and inspection metadata require no close operation and remain
+usable independently of the source files. A supplied `Reader` remains
+caller-owned and is never closed by ConfigKit.
+
+ConfigKit performs no background work after returning: it does not watch inputs,
+refresh secrets, or mutate published results. Inspection and rendering run only
+when explicitly called and use the retained snapshot. Applications normally load
+once near startup, but may explicitly load again using sources that satisfy their
+documented reuse rules. Additional loads produce independently owned results.
+
+Completion is not a transaction over external state. Files and environment
+variables are observed during processing rather than as one simultaneous external
+snapshot. ConfigKit does not roll back application callback effects or manage work
+or resources created by application callbacks. The explicit secrets-directory
+bootstrap remains a separate operation invoked by the application.
+
+## Layered loading
+
+`LoadLayered[T](main, options...)` accepts zero or one `LayeredOptions[T]` value.
+Its `main` is an `Input` containing a source, import opt-in, and optional
+filesystem context. A zero source is invalid; `Input{Source: NoSource()}` omits
+the main document. Each `Underlay` contains inputs that compose strictly as
+one tier, including each input's enabled imports. Empty groups contribute
+nothing. Underlay groups are listed from lowest to highest priority. Repeating
+a file within one group remains an import/repetition error; the same reusable
+source may be selected again in another tier.
+
+ConfigKit acquires and composes the main input first, constructs the schema,
+and checks the optional application version from the main tier only. Underlay
+roots cannot claim that version path (`underlay.version`); imported fragments
+retain `import.version` protection. It then acquires and composes underlay
+groups in order, recursively combines the groups, and applies the main tier,
+one selected overlay, and explicit CLI overrides. Scalar, sequence, and null
+definitions replace lower values between tiers; mappings combine recursively.
+Syntax and strict-composition errors inside a tier cannot be masked by a
+higher-priority tier. Defaults, typed decoding, resolution, secret lookup,
+validation, and publication run once on the final assembled input.
+
+An active layered overlay requires `LayeredOptions.OverlayFilesystem` explicitly,
+independent of the main source's namespace. Its name and root-relative lookup
+follow the selected-overlay rules below; inactive overlays ignore the context.
+The same path in a different filesystem namespace is not a repeated file.
+Ordinary `Load` continues to infer its selected-overlay context from its source.
+
+Source IDs follow acquisition order—main and its depth-first imports, then each
+underlay tier's inputs and depth-first imports, then overlay and CLI—even though
+history follows precedence order. Source `Layer` describes precedence while
+`Kind` describes acquisition. Underlay records use `LayerUnderlay` and a
+one-based `UnderlayIndex` giving their original tier position; all other records
+have index zero. Explicit empty groups and `NoSource` consume no source ID.
+Byte, source, node, and glob-entry budgets are shared across the load, while
+the existing per-source and import-depth bounds still apply.
+
+When a source mapping is explicitly placed as an item, its effective destination
+is the sequence container. After assembly assigns an index, nested source
+destinations are projected through that item. Physical source locations and
+resource contexts continue to identify the declaring source.
+
+#### Input destinations
+
+`Input.Destination` places the source's application mapping beneath a path
+relative to the configuration root. Its zero or empty value means root, so
+existing inputs keep their behavior. Main inputs and underlay inputs use the
+same destination model. Destination segments are literal mapping keys: dots,
+slashes, and numeric-looking text are not split or interpreted as sequence
+indexes. Explicit index segments and paths that introduce the reserved
+`configkit` or mapping-valued `extends` structure are invalid input options and
+produce a structured `LoadError` before source acquisition.
+
+Source-root `configkit.imports` directives are processed before placement. An
+import without an explicit destination inherits the effective destination of
+its importing document, including through nested imports. An import may instead
+declare a relative destination as a sequence of literal mapping keys; its keys
+append to the importing document's effective destination. Omission and `[]`
+add no keys. Nested imports continue relative to their immediate importing
+document, each declaration is independent of its siblings, and all matches of
+one glob receive the same resulting destination. This logical placement does
+not change how the physical import path resolves from its declaring file.
+
+A valid destination on `NoSource()` contributes nothing; an actual empty
+mapping materializes its destination containers. Once placed, contributions
+follow the same strict composition within a tier and recursive precedence
+behavior described above. A selected overlay and CLI overrides remain rooted at
+the application root and are not moved beneath the main input's destination.
+
+`SourceInfo.Destination` reports each acquired source's effective mapping or
+sequence-container path, including imports. `PlacementMode` reports how that
+source was placed; mapping is the zero value, and CLI/overlay sources report
+mapping unless item placement was applied. `DestinationUnresolved` is true only
+when an enclosing item index needed to name the source's destination has not yet
+been assigned. In that case `Destination` reports the nearest resolved enclosing
+collection container and omits the unresolved local suffix. A source's own item
+placement still has an exact destination at its sequence container. Completed
+loads project nested source destinations through their owning items. Callback and
+failed-load snapshots preserve the evidence available at that point; a fallback
+snapshot remains unresolved, while a later exact snapshot is false. Returned
+source records and paths are detached. Placement does not change source IDs,
+acquisition identity, physical source locations, or passive resource contexts.
+Introduced ancestor mappings use the contributing document's root location as a
+source fallback; authored content retains its original source coordinates.
+
+The maintained [layered consumer example](consumers/README.md#native-filesystem-and-layered-input)
+shows a main host file, two underlay tiers, embedded defaults, a mapping source,
+and explicit resource access without a second configuration pass.
+The accepted rationale for input placement and its relative-import extension
+is recorded in [ADR 0011](adr/0011-input-destinations.md) and
+[ADR 0012](adr/0012-relative-import-destinations.md).
+
+### Sibling inheritance
+
+Any input mapping may declare `extends: name` to inherit from the mapping named
+`name` among its immediate siblings. Names are exact keys, including the empty
+string; lookup does not search ancestor or unrelated scopes. The document root
+and a sequence-element mapping have no sibling namespace for their own string
+relationship, though mappings nested inside them can have one. An `extends`
+value of null cancels a relationship without looking up a parent. Other value
+types, missing or non-mapping parents, and cycles fail the load with value-free
+diagnostics.
+
+Within each sibling scope, direct mapping definitions resolve by dependency
+readiness. A parent is resolved before a child; among currently eligible
+definitions, the lexically smallest canonical semantic pointer is selected,
+including newly eligible definitions as soon as their parent completes. This
+ordering governs inheritance operations and their history, controls, and
+transfers. It does not reorder the mappings in the resulting configuration.
+
+`extends` is reserved structural input across ordinary and layered loads, every
+source kind, imports, selected overlays, and explicit CLI overrides. It is
+consumed before typed decoding and need not appear in the application's Go
+types. Applications previously using `extends` as an ordinary YAML field must
+rename that key and its field. No collection registration or keyword setup is
+required.
+
+ConfigKit completes strict composition within tiers and precedence assembly
+between tiers before resolving relationships. An omitted `extends` retains a
+lower-priority relationship; `extends: null` cancels it while preserving the
+child's own settings. Superseded references are historical evidence, not active
+relationships to validate. Each child recursively overlays its own settings
+over its resolved parent: explicit zero, false, empty string, sequence, and
+null values replace inherited values, while mappings combine recursively.
+The child retains its key. All effective relationships resolve, including
+entries the application may never select for execution. Partial parent
+definitions undergo normal typed decoding and application-owned validation,
+not a special structural exemption.
+
+Enclosing relationships resolve first. Nested instructions carried by a parent
+are then resolved separately at each destination, so a child's nested sibling
+override can change an inherited nested relationship. Typed defaults and
+application-derived values do not start another inheritance pass. Inherited
+input counts as explicit presence for `Metadata.IsExplicit`,
+`Resolution.IsExplicit`, and absence-based resolver defaults, even for false,
+zero, empty, or null input. `Origin.Inheritance` distinguishes it from a direct
+declaration with concrete source, destination, and controlling directive paths.
+A direct child replacement has its own effective origin; displaced ancestry
+remains inspectable in history. An inherited relative resource keeps the
+declaring source's passive `SourceContext`, not the child's directory.
+
+The optional schema-version checker still runs once against the main input at
+its original point. Inheritance cannot add, replace, or consume the protected
+version path; an unrelated relationship inside an enclosing mapping does not
+invalidate an unchanged version. Expansion uses the decoded projection
+node/depth limits without charging copied values as acquired sources;
+retained evidence uses the existing history-node limit. See the
+[consumer example](consumers/README.md#named-sibling-inheritance) for one-load
+usage.
+
+### Source resource contexts
+
+`Result.SourceContext(id)` and callback-scoped `Resolution.SourceContext(id)`
+return a passive `ResourceContext` for a source with filesystem context, or
+false for unknown and contextless IDs. The result query is nil-safe; the
+resolution query follows the callback's lifetime guard. Queries perform no
+filesystem access. Obtain IDs from exact origin locations rather than guessing
+one context for a composite origin or an application-derived value.
+
+`ResourceContext.FS` is rooted at the source's configured root, and `BaseDir`
+is its slash-relative directory inside that root. This includes imported
+sources, selected overlays, host files even without enabled imports, native
+files, and byte, reader, or mapping inputs with an explicit context. CLI and
+contextless synthetic inputs have no resource context. To open a relative
+resource, join `BaseDir` with the application resource name using slash-path
+operations and open that path through `FS`.
+
+For a resource-valued field, query `result.Metadata().Lookup(fieldPath)`, inspect
+its origin locations, choose a source ID only when the application's policy can
+identify the intended contributor, then call `result.SourceContext(id)`. The
+context belongs to that source even when the field survived from a lower-priority
+underlay. A composite origin has no automatically selected resource base.
+
+Native contexts borrow the caller's filesystem within the configured root.
+Host contexts retain a canonical root path, not an open root handle; each
+resource `Open` checks containment and opens a fresh handle that the returned
+file's `Close` releases. Applications own and close the files they open.
+Resources are application data, not configuration input, and do not consume
+configuration byte or node budgets. The loaded configuration and metadata stay
+usable if source files later change or disappear. Later resource operations
+observe the backing filesystem then and may fail if files move, disappear, or
+a caller-owned filesystem closes. This is not a snapshot or universal sandbox.
+
 ### Application schema versions
 
 `VersionRule` declares a semantic path and non-nil checker. Before defaults or typed
@@ -93,7 +322,7 @@ This is a load-local exception: siblings and unrelated fields, dynamic values,
 integer and float destinations, durations, `Secret`, and text codecs retain ordinary
 strict decoding. The retained source and history stay numeric while typed projection
 and effective YAML contain the resulting string; explanation records this through
-the normal decoding reconciliation event. Fragment, profile, and CLI ownership
+the normal decoding reconciliation event. Fragment, overlay, and CLI ownership
 protections apply before this conversion.
 
 The checker reports safe, value-free errors and warnings through `Validation.Error`
@@ -103,17 +332,79 @@ error without reporting one, ConfigKit emits a generic `version.hook_failed`
 diagnostic and suppresses the returned error text. Warnings alone are retained on a
 successful result. Callback panics remain programming failures.
 
+#### Shared source-version checks
+
+`Options.SourceVersionMode` and `LayeredOptions.SourceVersionMode` select how
+one `Definition.Version` policy applies across source documents. The zero value
+`SourceVersionModeDefault` preserves the existing main-only check at its
+effective path and existing protected-path behavior. `SourceVersionModeIfPresent`
+checks each participating document that declares a version and permits absent
+non-main declarations. `SourceVersionModeRequired` rejects a participating
+document without a declaration. Enabled modes require a usable `VersionRule`
+and checker; an unknown mode or unusable rule is an options error before source
+access or callbacks.
+
+In enabled modes, `Definition.Version.Path` is interpreted in each acquired
+document's own coordinates, before placement and composition. The main document,
+direct underlay documents, imported documents (including transitive and glob
+matches), and the selected overlay participate. `NoSource`, absent optional
+imports, empty underlay groups, defaults, resolution output, inheritance, and
+synthetic CLI overrides are not documents to check. An absent declaration in an
+actual main document still reaches the checker in if-present mode; required mode
+rejects it first. With `NoSource`, the existing single absent-main checker
+behavior remains in both enabled modes. If-present omissions in non-main
+documents skip the checker. Explicit null is present and reaches the checker;
+unsupported selected values produce `version.type`. No equality with the main
+declaration or non-null rule is imposed. Acceptance establishes only that the
+declaration is accepted by the consumer's checker; it does not establish that
+each fragment conforms to a complete application schema. Final assembled
+decoding and application validation still run once.
+
+A header that is absent below a structural control follows the same omission
+policy. Actual header/control overlap is rejected, and ordinary structural
+validation and main-version ownership protection continue independently.
+
+Accepted non-main declarations are consumed from their owned payload before
+composition; their authored ancestors remain, even if empty. The main
+declaration remains in the application. Ordinary release metadata should use a
+different path. Enabled main protection uses the main input's effective path:
+its `Destination` prefixed to the source-local `VersionRule.Path`. This protects
+the remaining payload against imports, underlays, inheritance, and CLI
+overrides. Thus a mounted main looks up a local header while ownership checks
+protect its destination-prefixed application path. Main checker diagnostics and
+declaration locations use source-local coordinates; ordinary payload
+diagnostics continue to use effective paths. Default-mode diagnostics are
+unchanged.
+
+`SourceInfo.VersionCheck` is nil when enabled declaration processing did not
+occur. Otherwise it records a source-local `Path`, the observed `Version` state
+(`Present`, `Null`, and, when disclosed, exact `Text`), an actual or nearest
+available `Location`, and whether text was redacted. A non-nil record with
+`Present == false` records observed absence; it does not say the checker ran or
+accepted a declaration. Absent evidence still identifies the full requested
+header path, while its location identifies the nearest available source node.
+During callbacks and in failed-load source records,
+text is conservatively withheld because final classification may not be
+available. Successful final source records reflect completed sensitivity
+classification at the declaration's mounted path and ancestors, including
+historical classification. Redaction empties `Text` while retaining presence
+and null state. Returned copies are detached; evidence does not reopen inputs.
+Consult [ADR 0013](adr/0013-shared-source-version-checks.md) for the accepted
+decision and the [development guide](development.md#baseline-validation) for
+the completed implementation and remediation records in Git history.
+
 ### Layered metadata
 
 `Result.Metadata` describes every final canonical path. Queries retain explicit
 key/index distinctions using actual effective and historical container shapes;
 only parsed numeric tokens are resolved contextually. A descendant that never
-existed is not historical merely because an atomic ancestor was replaced. Exact surviving input paths
-are explicit and have a base or profile origin; omitted paths have a default origin.
+existed is not historical merely because an atomic ancestor was replaced. Exact
+surviving direct or inherited input paths are explicit and retain their source
+layer and derivation; omitted paths have a default origin.
 An explicit parent mapping does not make omitted children explicit. Null state and
 available source coordinates are retained. `History` records the Go-zero or application
-default baseline followed by exact base or profile definitions, including
-equal-valued input. Profile replacement removes displaced descendants from effective
+default baseline followed by exact base or overlay definitions, including
+equal-valued input. Overlay replacement removes displaced descendants from effective
 metadata while retaining their base definition and the parent replacement operation
 in history. All metadata queries and returned nested records are detached.
 
@@ -124,22 +415,38 @@ domain ownership rules remain application responsibilities. The method's Go
 documentation in [metadata.go](../metadata.go) defines enumeration and path-ordering
 details.
 
-History is chronological: a path can have a synthetic default record, its exact base
-definition, a profile replacement, and a recorded resolution operation. A descendant
+History is chronological: a path can have a synthetic default record, its exact
+base definition, an overlay replacement, structural inheritance, and a recorded
+resolution operation. A descendant
 query uses the ancestor's path for a whole-container or whole-sequence replacement;
 the descendant can therefore retain history after `Lookup` stops reporting it as
 effective. Equal-valued input is still an explicit base definition, while a derived
 value for absent input remains non-explicit.
 
+Appending registered records during strict composition assigns each contribution
+its assembled sequence index in history. A later precedence tier that supplies a
+sequence records replacement at the sequence path; history does not imply
+identity-based merging across tiers.
+
 A mapping assembled from the selected file and imports has one composite base origin
-whose locations retain source traversal order. A mapping also visited by a profile
-has a profile-layer composite origin with the lower contributing locations followed
-by the profile location; an empty mapping is still an explicit container even though
-it contributes no children. Resolution origins are synthetic: they have layer
+whose locations follow [scoped assembly order](#explicit-imports), not source-record
+acquisition order. A mapping also visited by an overlay has an overlay-layer
+composite origin with the lower contributing locations followed by the overlay
+location; an empty mapping is still an explicit container even though it contributes
+no children. Resolution origins are synthetic: they have layer
 `resolution`, no fabricated file location, and a `derived_default` or `normalization`
 operation. Reasons and dependency paths retain first occurrence order with duplicates
-removed. Source records remain ordered as selected source, depth-first imports, the
-selected profile, then synthetic CLI records in supplied order.
+removed. Source records and IDs retain [acquisition order](#explicit-imports),
+including the underlay ordering described in [layered loading](#layered-loading).
+
+Inheritance adds no acquired source or precedence layer. `Origin.Inheritance`
+lists concrete transfers from contributor paths through destinations, with
+each step's controlling `extends` path and location. `History` retains source
+applications at their original paths and projects parent contributions, child
+replacements, and consumed controls at destination paths after source assembly
+and before typed decoding. `HistoricalPaths` includes consumed directives and
+displaced inherited descendants. A control can have history even when `Lookup`
+reports it absent. Queries return detached records.
 
 ### Secret classification and effective rendering
 
@@ -156,6 +463,11 @@ Propagation reaches a fixed point, so record/map iteration order does not affect
 result and dependency cycles terminate. A dependency on a classified descendant or
 on a container that contains one is treated conservatively as sensitive. Applications
 must classify secret-producing transformations whose dependencies are not recorded.
+Inherited value and control paths link source and destination classifications
+through chains and nested copies; unrelated sibling fields are not classified
+merely because another field is secret. Displaced historical evidence remains
+redacted. Secret references are copied only as unresolved names; credential
+lookup stays after decoding and application resolution.
 
 `Result.WriteEffective` writes the frozen resolved snapshot as deterministic redacted
 YAML. Mappings sort by decoded key string, sequences retain order, indentation is two
@@ -188,9 +500,14 @@ historical path. Its effective value and each chronological before/after value a
 `ValueSnapshot` records containing deterministic YAML text rather than mutable Go or
 parser values. `Present=false` with empty YAML means absence or deletion; explicit
 null is present and renders as `null`. The history starts with the canonical typed
-application-default or Go-zero baseline, then includes only relevant base sources in
-traversal order, the selected profile, supplied CLI records, and audited net
-resolution operations. Overlapping parent/child resolution records are rendered
+application-default or Go-zero baseline, followed by relevant underlay and main-tier
+contributions in precedence order, the selected overlay, supplied CLI records,
+and audited net resolution operations. Within each strict tier, replay follows
+[scoped assembly order](#explicit-imports), not source-record acquisition order.
+Structural inheritance events appear after source replay
+and before typed reconciliation, including consumed or canceled `extends` input
+and equal-valued child replacements. Overlapping parent/child resolution records
+are rendered
 as one net subtree event with combined reasons and dependencies, rather than
 pretending that the recorded assignments ran in lexical order. Explicit equal
 writes remain visible. When input replay
@@ -199,7 +516,7 @@ before domain resolution. Its origin retains the decoded path's source/default
 attribution; its reason identifies static-default retention and canonical decoding.
 This built-in inspection operation is not accepted by `Resolution.Record`.
 
-Base, profile, and CLI history preserves the policy-validated input representation,
+Base, overlay, and CLI history preserves the policy-validated input representation,
 including a superseded value that could not have decoded into the final Go field.
 Baseline, effective, and resolution values use canonical typed projection. A
 decoding event bridges replayed input and canonical typed values, so intermediate
@@ -238,6 +555,46 @@ an explicit `LoadSecretsDir` bootstrap, then use `WriteDiagnostics`,
 ConfigKit does not infer bootstrap outcomes from a secret reference and does not
 bundle these independent reports into a command framework.
 
+## Effective configuration diffs
+
+`configkit.Diff(left, right)` compares two completed successful loads of the same
+configuration type. `left` is the earlier state and `right` is the later state.
+Applications load both sides with the sources and options they choose; to inspect
+an overlay's impact, load the same base once without the overlay and once with it
+selected. The [consumer guide](consumers/README.md#effective-configuration-diffs)
+links to maintained examples. The public API returns safe changed-region records;
+applications own their presentation.
+
+Comparison uses the frozen canonical effective inspection values, not mutable
+`Result.Config`, source text, provenance, or history. It does not reopen sources,
+reread environment variables, or rerun defaults or callbacks. Equal effective
+values produce no records even when sources, explicit-input status, defaults, or
+resolution histories differ; for example, explicit zero and defaulted zero are
+equal. Mappings compare recursively, sequences compare atomically, and scalar,
+kind, or null/container replacements report their affected region. Records use
+the existing comparator's canonical pointer ordering. `Present=false` with empty
+YAML means absence; explicit null is present and renders as `null`.
+
+Before either side is returned, sensitivity from both loads is applied to both
+snapshots, including historical and dependency-derived classification. A
+classification on either side protects the corresponding values on both sides.
+`Redacted` reports masking, while `HiddenChanged` reports an actual concealed
+value or presence change; it is not implied merely by redaction. A structural
+replacement that prevents safe correspondence masks the complete reported
+region, and an absent side remains absent. Record membership establishes that a
+change occurred even when safe snapshots render identically. Literal
+marker-looking input is ordinary input; consumers use the record fields rather
+than interpreting marker text.
+
+Typed secret references contribute their configured environment names to the
+frozen comparison, never resolved credential bytes. Credential rotation behind
+an unchanged reference is therefore not detected. The records contain changed
+regions only, not paired full-document context; independently calling
+`WriteEffective` on each result does not create cross-protected context. Nil or
+uninitialized results return an error with no records. A failed `Load` retains
+its normal nil-result/`LoadError` behavior, which callers handle before invoking
+`Diff`.
+
 ## Typed secret references
 
 Credential-bearing fields use the opaque `Secret` type and store an environment
@@ -338,7 +695,7 @@ found entry path. Resolution changes, additions, removals, and moves do not alte
 this entry view.
 
 `Resolution.Source(id)` returns a detached record for an already acquired source in
-the same load: the main source, imports, selected profile, and supplied CLI records
+the same load: the main source, imports, selected overlay, and supplied CLI records
 where applicable. Empty and unknown IDs return false. It never reloads input, reads
 the filesystem, resolves symlinks, or looks up credentials. A source path exists
 only when the acquired source has one; defaults and byte, reader, and CLI sources do
@@ -421,56 +778,91 @@ Application defaults do not participate in this operation.
 | Mapping | Mapping | Recurse by mapping key |
 | Any other pair of present nodes | Any other pair of present nodes | `compose.conflict` |
 
-Every non-mapping overlap is a conflict, even when values are equal. Nulls and
-sequences are present atomic values, so they also conflict when repeated; sequences
-are never joined by index. Disjoint entries under shared mappings compose, making
+Every non-mapping overlap is a conflict, even when values are equal, except for
+sequence/sequence contributions at a registered sequence-composition path. Nulls
+remain atomic conflicts. Unregistered sequences conflict when repeated and are
+never joined by index. Disjoint entries under shared mappings compose, making
 stable keyed mappings the natural representation for independently contributed
 entries.
 
+`Definition.SequenceCompositions` registers literal mapping paths and one literal
+`IdentityField` for sequences that should compose independently authored records.
+Registration paths follow input-destination key rules: parsed numeric tokens are
+literal mapping keys, explicit index segments and reserved `configkit` or
+mapping-valued `extends` segments are invalid, and the root registration is
+harmless and unused. The identity field is one literal string key; neither the
+field name nor identity strings are normalized.
+At each registered path, every actual sequence contribution in a strict tier is
+validated, including a singleton contribution: each record must be a mapping with
+a present string identity field. Distinct exact identity strings append in source
+and list order; duplicate identities conflict even if the records are otherwise
+equal or disjoint. Empty identity strings are valid. Malformed records produce
+`compose.invalid`; duplicate identities produce `compose.conflict`, with the
+sequence container and source locations identifying a duplicate without exposing
+identity values. Missing paths and non-sequence values do not fail by themselves.
+Identical registrations coalesce; conflicting registrations at one path and
+invalid registration paths produce `options.invalid`. Unused and root registrations
+are allowed and have no effect.
+
+Registration changes strict composition only. Each underlay group and the main
+tier has independent duplicate-identity state. Precedence assembly is unchanged:
+a present higher-priority sequence replaces the lower sequence as one value,
+including at a registered path. Overlays, CLI overrides, defaults, inheritance,
+and application resolution do not run the registered identity check. The accepted
+rationale is recorded in [ADR 0014](adr/0014-sequence-placement-and-composition.md).
+The [maintained prompt catalog](consumers/README.md#identified-prompt-sequences)
+shows item path and glob imports, helper-supplied identity, and main-tier
+replacement of a lower-priority sequence.
+When an enabled source-version mode is selected, declarations are checked in each
+source's own coordinates before placement and composition. Accepted non-main
+headers are consumed before strict composition and cannot supply a registered
+record identity. Final typed decoding and application validation still run on the
+completed configuration.
+
 An empty mapping contributes no child definitions. Two empty mappings compose, and
 an empty mapping neither clears nor conflicts with another mapping's children.
 Shared mapping containers retain all contributing value locations as their composite
-base origin. A conflict reports the first claim as its primary location and every
-remaining claim as related locations, in deterministic source and declaration
-traversal order. Traversal order never authorizes replacement. Composition clones
-the resulting tree and its ordered claim records rather than mutating or aliasing an
-input document.
+base origin. A conflict reports the first claim in
+[scoped assembly order](#explicit-imports) as its primary location and every
+remaining claim as related locations in the same order. Assembly order never
+authorizes replacement. Composition clones the resulting tree and its ordered
+claim records rather than mutating or aliasing an input document.
 
-## Selected profiles
+## Selected overlays
 
-`Options.Profile` selects at most one overlay by name. Names must match
+`Options.Overlay` selects at most one overlay by name. Names must match
 `[a-zA-Z0-9_-]+`; they are not paths. The empty name and exact lowercase name
-`default` select no profile, and `profiles/default.yml` is never loaded implicitly.
-Every other name resolves only to `<configuration-root>/profiles/<name>.yml`, with
-no extension fallback, case normalization, inheritance, or profile stack. A selected
-missing profile is an error.
+`default` select no overlay, and `overlays/default.yml` is never loaded implicitly.
+Every other name resolves only to `<configuration-root>/overlays/<name>.yml`, with
+no extension fallback, case normalization, inheritance between overlay files,
+or overlay stack. A selected missing overlay is an error.
 
-The profile is a separately policy-validated YAML mapping. It is loaded after the
+The overlay is a separately policy-validated YAML mapping. It is loaded after the
 base import graph is composed and its schema version is checked, but before defaults
-and typed decoding. Profile bytes, nodes, and source records count toward the same
+and typed decoding. Overlay bytes, nodes, and source records count toward the same
 aggregate limits as base inputs. Its canonical path must remain within the configured
 root and must not duplicate a file already present in the base graph.
 
 Overlay is intentionally distinct from strict composition:
 
-| Existing lower node | Incoming profile node | Result |
+| Existing lower node | Incoming overlay node | Result |
 | --- | --- | --- |
-| Absent | Value | Retain the profile value |
+| Absent | Value | Retain the overlay value |
 | Value | Absent | Retain the lower value |
 | Mapping | Mapping | Recurse by mapping key |
 | Any other pair of present nodes | Any other pair of present nodes | Replace the entire lower node |
 
-Scalars, sequences, and null are atomic. Thus a profile sequence replaces rather
+Scalars, sequences, and null are atomic. Thus an overlay sequence replaces rather
 than extends a base sequence, `[]` clears it, null is an explicit replacement, and
 cross-kind replacement is allowed when the final application schema accepts it.
 Replacing a subtree removes all displaced descendants from effective presence and
 origin data; history retains the displaced definitions and identifies the parent
-replacement. Surviving profile paths have profile origins, while recursively shared
+replacement. Surviving overlay paths have overlay origins, while recursively shared
 mapping containers retain their contributing locations.
 
-A profile may not contain `configkit` anywhere, contribute the configured schema
+An overlay may not contain `configkit` anywhere, contribute the configured schema
 version, or replace an ancestor of that version with a scalar or null. These rules
-apply even if the supplied version text would be unchanged. Profiles cannot mask a
+apply even if the supplied version text would be unchanged. Overlays cannot mask a
 base parse, composition, or version failure because those checks run first.
 
 ## Explicit CLI overrides
@@ -483,8 +875,8 @@ must never contain the raw argument or configured value. General environment
 variable overlays are an explicit non-goal: environment access is reserved for
 typed secret references.
 
-Overrides apply after the selected profile and before typed defaults/decoding, so
-the input precedence is application defaults < composed base < selected profile <
+Overrides apply after the selected overlay and before typed defaults/decoding, so
+the input precedence is application defaults < composed base < selected overlay <
 CLI. Recorded resolution, typed secret lookup, and semantic validation follow the
 effective override tree. Overrides also work with `NoSource`.
 
@@ -506,7 +898,7 @@ configured schema-version path. Unknown fields and effective type mismatches rem
 ordinary strict-decoder errors.
 
 At most 1,024 records are accepted, and each record creates one synthetic CLI
-source after file/profile sources in supplied order. Override sources share the
+source after file/overlay sources in supplied order. Override sources share the
 1,024-source ceiling. A record has an 8 MiB string-byte limit; override strings and
 source bytes share the 64 MiB aggregate limit. Synthetic mapping keys and values
 count toward the existing 1,000,000-node and depth-256 ceilings. Cycles, named or
@@ -514,7 +906,7 @@ unsupported types, nonfinite floats, and limit excesses fail with safe `override
 diagnostics. Diagnostics and provenance identify the safe `Flag` source, never raw
 arguments or values.
 
-Override origins are explicit CLI origins. History retains replaced base/profile
+Override origins are explicit CLI origins. History retains replaced base/overlay
 definitions and records equal-value replacements as well as structural changes.
 Caller-owned maps and slices are snapshotted during loading and must not be mutated
 concurrently with the call. The executable `ExampleOverride` shows translating a
@@ -525,10 +917,18 @@ zero.
 
 Imports are opt-in. With `Options.Imports` enabled, a document may reserve its root
 `configkit` mapping for an `imports` sequence. ConfigKit removes that mapping before
-application decoding. Each sequence item is a mapping with exactly one nonempty
-string `path` or `glob` plus an optional boolean `optional`; all other shapes and
-fields are `import.invalid`. A `configkit` key below the root is invalid. With
-imports disabled, any occurrence remains `import.disabled`.
+application decoding. Each sequence item is a mapping with exactly one `path`,
+`dir`, or `glob` selector. `path` and `dir` accept a string or a list of strings;
+`glob` accepts one path pattern string. The optional `required` boolean controls
+whether an empty selection or a missing named path is an error. Its default is
+false, so missing imports are absent unless `required: true` is specified. The
+removed `optional` field is invalid. `destination` is an optional sequence of
+strings. Every destination string is one literal mapping key and follows the
+input-destination validation rules above; quote numeric-looking keys so YAML
+reads them as strings. Null, scalar destinations, and non-string members are
+`import.invalid`, as are all other unsupported shapes and fields. A `configkit`
+key below the root is invalid. With imports disabled, any occurrence remains
+`import.disabled`.
 
 The directive key is `configkit`, not the `_configkit` spelling used in v0.1.0.
 There is no legacy alias: `_configkit` is ordinary application data, subject to
@@ -537,38 +937,132 @@ directive keys in root files and imported fragments, and rename any application
 fields previously named `configkit` to avoid the newly reserved key.
 
 Path declarations use slash-separated relative paths. Absolute paths and network
-URLs are invalid. Each path resolves from its declaring file's canonical directory,
-and files are traversed depth-first in declaration order: a file contributes its own
-body before its imports. Order stabilizes source records and diagnostics but never
-permits replacement. Required missing paths fail; `optional: true` suppresses only
-target nonexistence, not invalid paths, containment failures, directories,
-permissions, parse errors, or composition errors.
+URLs are invalid. Each path resolves from its declaring file's canonical directory.
+An empty `path: []` selects nothing; it succeeds even when `required: true` because
+there are no named targets to be missing. A nonempty required path list verifies
+every named target, including targets later excluded from selection.
+
+Directory declarations search the listed literal directory roots. Missing roots
+contribute no candidates; supplied inaccessible or invalid roots fail. `pattern`
+accepts one basename pattern or a list, matching filenames only. Omission selects
+all filenames; an explicit empty list selects none. `exclude` accepts one basename
+pattern or a list and removes matching final candidates. Patterns use Go's
+case-sensitive `path.Match` grammar. `recursive` enables descendant traversal;
+`depth` is valid only when recursion is effective, with direct files at depth 1 and
+depth 0 or omission meaning unlimited traversal subject to library budgets.
+`follow` controls discovered symlink entries and `hidden` controls discovered
+names beginning with a dot. Both default to true. Explicit roots retain their
+normal access behavior.
 
 Glob declarations use the slash-based Go `path.Match` grammar: `*`, `?`, character
-classes, and backslash escapes are supported, while recursive `**` syntax is invalid.
-Backslashes escape pattern characters rather than acting as directory separators.
-ConfigKit validates the complete pattern before filesystem traversal, walks only
-directories selected by preceding segments, and reads wildcard directories in
-bounded batches. Every examined entry consumes the separate visit budget, including
-nonmatches. Matches are sorted lexicographically by slash-relative spelling within
-each declaration before depth-first source traversal. A required zero-match glob is
-`import.missing`; `optional: true` suppresses only zero matches. Invalid patterns,
-walk failures, matched directories, invalid documents, escapes, and repeated targets
-remain errors. Final candidates are bounded by the remaining source limit before
-their contents are accumulated.
+classes, and backslash escapes are supported, while recursive `**` syntax is
+invalid. Backslashes escape pattern characters rather than acting as directory
+separators. ConfigKit validates the complete pattern before filesystem traversal,
+walks only directories selected by preceding segments, and reads wildcard
+directories in bounded batches. Every examined entry consumes the separate visit
+budget, including nonmatches. Literal components are explicit path components;
+`follow` and `hidden` filter only wildcard-discovered entries. `exclude` filters
+final candidate basenames.
 
-The configuration root defaults to a selected file's canonical directory. A file
-source may supply `FilesystemContext.Root` to widen that root but may not supply
-`BaseDir`. Whenever byte and reader sources supply filesystem context, they must
-provide both `BaseDir` and `Root` as absolute existing directories; their diagnostic
-labels, including URI-looking labels, never imply filesystem paths. The base and
-selected file must be inside the canonical root. Nonregular targets are rejected before opening to avoid blocking on FIFOs;
+`Options.ImportDefaults` and `LayeredOptions.ImportDefaults` provide one
+application-owned load-wide policy for enabled imports in the main source,
+underlays, and transitive imports; they do not apply to selected overlays or
+discover sources by themselves. Each applicable field resolves independently:
+an explicit declaration value wins, then the application default, then the
+library default (`required: false`, `follow: true`, `hidden: true`,
+`recursive: false`, and unrestricted filenames). A declaration's `pattern`
+replaces rather than extends the default pattern. Nil `Pattern` leaves directory
+filenames unrestricted; a non-nil empty slice selects none. Defaults do not make
+a field applicable to selectors that do not support it, and load calls snapshot
+the supplied pointers and slice before source acquisition.
+
+Acquisition visits declarations in authored order and each declaration's selected
+paths in namespace-path lexical order, preserving source-record and version-callback
+order. Ordinary mapping contributions in a main tier are assembled dependency-first:
+ConfigKit selects the lexically smallest stable source key among the contributions
+whose imports are complete. Independent branches may interleave, and the main body
+may be assembled after its imports. Sequence-item contributions in that tier are
+nodes in the same schedule; each selected item first assembles its own strict
+mapping scope, with helpers before the bodies that import them. Those local scopes
+do not use application collection policies and are not flattened into the outer
+schedule. Item positions and diagnostic paths follow the resulting assembly order.
+Each underlay group likewise acquires its inputs before dependency-ready assembly
+in one strict scope. Independent groups remain separate precedence tiers; a higher
+tier still replaces a lower tier's sequence values wholesale. Assembly order never
+permits replacement.
+
+Acquisition order and resolution order serve different purposes: acquisition
+assigns source records and callback order, while the shared scheduler determines
+composition order among currently dependency-ready contributions. Inheritance
+uses the same eligibility rule within each destination-local sibling scope; it
+does not reorder the authored output mapping. The private namespace and stable-key
+construction is described in the [import implementation overview](internal/imports.md).
+
+An empty selection is allowed by default. `required: true` makes an empty directory
+or glob selection fail with `import.missing`; it also makes missing named paths
+fail. Requiredness does not suppress invalid paths, containment failures,
+directories selected as files, permissions, parse errors, or composition errors.
+
+For inspection, `SourceInfo.ImportSelections` records one outcome for each
+completed optional declaration whose entire selected set is empty. The record is
+attached to its declaring source, keeps the declaration path local to that source,
+identifies the declaration mapping (including its `SourceID`) and selector, and
+uses outcome `none`. Records remain in declaration order and are available
+through the existing result, load
+error, and callback source snapshots. They do not enumerate candidates, explain
+why a path was absent, or describe partial omissions. Empty `path: []` succeeds
+regardless of requiredness: it produces a `none` record when effectively optional,
+and no record when effectively required. Snapshots detach the record paths and
+slices.
+
+Each declaration may include `placement: mapping` or
+`placement: sequence-item`. Omission means `mapping` for that declaration,
+including every match of a glob; placement is not inherited. Sequence-item
+placement requires a nonempty effective destination and wraps the imported
+mapping as one element of a sequence there. It does not split source sequences
+or infer identity fields. Repeated sequences still conflict unless their path is
+registered through `Definition.SequenceCompositions`; registration does not
+change placement or precedence behavior. Sequence records retain source
+locations through placement and history; indexes describe their assembled
+positions while source coordinates continue to identify the declaring files.
+Header/version policy and final application validation retain their existing
+owners and timing; sequence registration is a strict-composition rule only.
+
+An item import first opens a local mapping scope. Its body and ordinary helper
+imports compose within that record. An omitted helper destination or
+`destination: []` places the helper at the record root; other destinations are
+relative to the immediate importing source. A helper may supply fields such as
+the record's `id`. An explicit nested `sequence-item` import creates a sequence
+inside the record, and its placement is independent of the enclosing item.
+At the record root, an item import must name a nonempty nested destination.
+Physical paths continue to resolve from the file that declares each import.
+An actual empty mapping contributes an empty item, while an unmatched declaration
+with default or explicit `required: false` contributes nothing.
+
+For host files, the configuration root defaults to the selected file's canonical
+directory. A host file may supply `FilesystemContext.Root` to widen that root but
+may not supply `BaseDir` or a native `FS`. Byte, reader, and mapping inputs with host context
+must provide both `BaseDir` and `Root` as absolute existing directories.
+For `Filesystem.File`, the default root is `.` in its namespace and the base is
+the selected file's directory. An explicit context must use the same `Filesystem`
+namespace and may set `Root` to contain the file and its base; it may not supply
+`BaseDir`. Byte, reader, and mapping inputs with native context provide both `BaseDir` and
+`Root` as slash-relative `fs.ValidPath` directories in `FS` (`.` means its root),
+with the base inside the root. Context paths are checked even when imports and
+overlays are inactive. Diagnostic labels, including URI-looking labels, never
+imply filesystem paths. Nonregular targets are rejected before opening to avoid blocking on FIFOs;
 post-open regular-file checks remain in place under the stable-input assumption.
 ConfigKit canonicalizes symlinks,
 checks containment with relative paths, and opens the canonical root-relative target
 through a library-owned root handle. Contained relative or absolute symlink targets
 are accepted; escapes are rejected. Canonical paths define identity, so hard links at
-distinct paths remain distinct sources.
+distinct paths remain distinct sources. Native filesystems use slash-path
+operations. Where the supplied `fs.FS` implements `fs.ReadLinkFS`, ConfigKit uses
+that capability to inspect and resolve discovered links; otherwise it follows
+the filesystem's ordinary `fs.FS` behavior. ConfigKit does not require every
+native filesystem to expose link inspection, and native filesystem behavior is
+not a general sandbox guarantee. Imports stay in their selected namespace and
+root.
 
 An active-stack repeat is `import.cycle` and reports the complete cycle's import
 sites. A previously visited canonical path is `import.repeated` and reports both

docs/consumers/README.md

@@ -1,13 +1,13 @@
 # Consumer Adoption Guide
 
 ConfigKit removes parser setup, strict decoding, presence bookkeeping, import and
-profile traversal, precedence, provenance, secret lookup, and safe inspection from
+selected-overlay traversal, precedence, provenance, secret lookup, and safe inspection from
 applications. Consumers still own their configuration structs, defaults, semantic
 validation, schema-version policy, CLI parsing, runtime matching and conversion,
 resource checks, remote acquisition, and service initialization.
 
 The maintained [`examples/consumers`](../../examples/consumers) package demonstrates
-four sanitized shapes using only ConfigKit and the Go standard library:
+sanitized shapes using only ConfigKit and the Go standard library:
 
 | Shape | Demonstrated fit | Application code deliberately retained |
 | --- | --- | --- |
@@ -15,6 +15,13 @@ four sanitized shapes using only ConfigKit and the Go standard library:
 | Notarius-like | Canonical object bindings, a string-backed schema version accepting string or integer YAML, strict dynamic options, and `int64` dynamic integers | Catalog selection, allowed option keys, range-checked conversion to runtime `int` |
 | WeatherReporter-like | Defaults-only loading, report maps, canonical module objects, typed option unions, explicit CLI records, and typed secret startup | Report/module compatibility, rejection of explicitly empty domain lists, CLI parsing, and startup ordering |
 | Seriatim-like | Independent strict ordered speaker and autocorrect rule files | First-match and ordered replacement algorithms; no central application configuration file |
+| Bundled file | Simple native `fs.FS` source through ordinary `Load` | Selection and lifetime of the application-owned embedded filesystem |
+| Layered catalog | Embedded defaults with an import, a mapping underlay, an independent host main file, explicit false/zero input, and source-relative resource access | Domain validation and opening/interpreting the template resource |
+| Mounted collections | Entry-declared relative destinations, nested metadata/prompt/profile imports, local inheritance, and a version-only higher-tier override | Collection IDs, profile meaning, and opening/interpreting prompt resources |
+| Named presets | One-load sibling inheritance, explicit false, nested destination resolution, and source/derivation inspection | Choosing and executing a preset |
+| Effective configuration diff | Same-type frozen effective comparisons for two selected files or base-versus-overlay impact | Source/options selection and consumer-owned safe presentation |
+| Shared source-version checks | One `Definition.Version` policy for accepted declarations across participating source documents, with optional or required declarations and sanitized evidence | Choosing which declarations the application accepts; final assembled schema and domain validation |
+| Identified prompt sequences | Explicit item path/glob imports, helper-supplied IDs, reusable single-field registration, and main-tier replacement of a lower-priority collection | Prompt meaning and final domain validation |
 
 The YAML under [`testdata`](../../examples/consumers/testdata) is loaded by the
 executable examples and ordinary tests. It contains no private application data,
@@ -34,7 +41,7 @@ features appear only when the application needs them:
 result, err := configkit.Load[Config](configkit.File(path), configkit.Options[Config]{
     Definition: rules,
     Imports:    useDeclaredImports,
-    Profile:    selectedConfigProfile,
+    Overlay:    selectedConfigOverlay,
     Overrides:  explicitlySetFlags,
 })
 if err != nil {
@@ -43,11 +50,171 @@ if err != nil {
 return start(result.Config)
 ```
 
-Simple consumers use neither imports nor profiles. A complex consumer can opt into
-both without redefining their syntax. ConfigKit profiles are configuration overlays;
+Simple consumers use neither imports nor selected overlays. A complex consumer can opt into
+both without redefining their syntax. ConfigKit overlays are configuration overlays;
 they do not select or replace an application's LLM, model, pipeline, report, or
 execution profile concepts.
 
+## Import defaults and selection inspection
+
+`Options.ImportDefaults` and `LayeredOptions.ImportDefaults` apply one
+load-wide policy to enabled main, underlay, and transitive imports. Each field
+is resolved independently: a declaration's explicit value wins, then the
+application default, then the library default. `Required`, `Follow`, `Hidden`,
+and `Recursive` use nil pointers for unspecified values. `Pattern` applies to
+directory searches: nil leaves filenames unrestricted, while a non-nil empty
+slice selects no files. A declaration's own `pattern` replaces the application
+default. The defaults do not enable imports or change selected-overlay behavior.
+
+`Result.Sources`, `LoadError.Sources`, and `Resolution.Source` expose each
+declaring source's `ImportSelections`. These records identify the source-local
+declaration path, its authored location and selector, and the `none` outcome;
+they do not include candidate names or a guessed reason for absence. Returned
+records are detached snapshots. The [configuration contract](../config.md#explicit-imports)
+owns selector applicability and exact absence semantics.
+
+## Shared source-version checks
+
+[`LoadVersionedService`](../../examples/consumers/source_version.go) demonstrates
+if-present and required modes with one `Definition.Version` rule across a mounted
+main input and an underlay. The example reads only sanitized
+`SourceInfo.VersionCheck` fields; its executable test is in
+[`layered_test.go`](../../examples/consumers/layered_test.go). The
+[configuration contract](../config.md#shared-source-version-checks) defines
+participation, source-local lookup and diagnostics, non-main header consumption,
+main-path protection, default behavior, and evidence disclosure. Accepted
+declarations do not replace final assembled decoding or application validation.
+
+## Effective configuration diffs
+
+[`CompareFiles`](../../examples/consumers/diff.go) loads two caller-selected files
+with one application definition, handles each load error, calls `configkit.Diff`,
+and presents the safe records through consumer-owned `FormatDifferences`.
+[`CompareRolloutOverlay`](../../examples/consumers/diff.go) loads the same base
+once without an overlay and once with the selected overlay, so omitted lower
+priority values remain present. Its resolver records a derived endpoint URL after
+overlay application, and its nested typed secret reference is classified before
+the resulting records are presented.
+
+The maintained [behavior tests and executable examples](../../examples/consumers/diff_test.go)
+use temporary files and disposable secret values. Presenters should use
+`ValueSnapshot.Present` for absence, record membership for change, and
+`HiddenChanged` for a withheld change indicator. Do not parse the redaction marker
+or add separately rendered full contexts that lack combined sensitivity. The
+[configuration contract](../config.md#effective-configuration-diffs) owns shared
+equality, granularity, and disclosure semantics.
+
+## Named sibling inheritance
+
+[`LoadPresets`](../../examples/consumers/inheritance.go) loads the maintained
+[`presets.yml`](../../examples/consumers/testdata/inheritance/presets.yml) in one
+call. `custom` inherits `enabled: false` from `base`, while its nested
+`settings.primary` changes the inherited `settings.secondary` result at the
+custom destination. The application struct has no `extends` field; ConfigKit
+consumes the relationship before strict typed decoding.
+
+The [executable example](../../examples/consumers/consumers_test.go) reads
+`Metadata.IsExplicit` for the inherited false value, then follows
+`Origin.Inheritance` back to the parent path while `Origin.Locations` identifies
+the declaring source. Applications still choose which preset to execute and
+validate domain meaning. The [configuration contract](../config.md#sibling-inheritance)
+owns exact lookup, precedence, cancellation, and migration rules.
+
+## Native filesystem and layered input
+
+[`LoadBundled`](../../examples/consumers/layered.go) keeps the simple `Load` shape:
+it selects a file in an application-owned embedded filesystem and needs no
+underlay setup. [`LoadCatalog`](../../examples/consumers/layered.go) uses
+`LoadLayered` with an imported embedded-defaults group, a separate
+`Mapping` group, and a host main file supplied by its caller. The groups
+compose strictly within themselves and recursively across precedence boundaries;
+the main file's explicit `false` and `0` replace lower values without being
+mistaken for omissions. Its validator owns allowed modes and resource-name policy.
+
+[`ReadWelcome`](../../examples/consumers/layered.go) follows the effective field's
+origin location to a source ID, queries `Result.SourceContext`, and opens the
+resource relative to that context's `BaseDir`. It does not assume the main file
+owns a retained underlay value. A composite origin needs an application choice;
+ConfigKit does not guess a directory. The context query is passive, while later
+resource reads observe the caller-owned filesystem or a freshly opened host root
+and may fail independently of the completed configuration load. The
+[configuration contract](../config.md#source-resource-contexts) owns exact
+lifetime and precedence rules.
+
+[`LoadPromptBundle`](../../examples/consumers/layered.go) is the ordinary-root
+counterpart: its entry opts into imports through `Load`, imports metadata at the
+root, and places a prompt beneath `prompts` using YAML. The executable example
+and behavior test are in
+[`layered_test.go`](../../examples/consumers/layered_test.go).
+
+## Input destinations and mounted collections
+
+Main and underlay inputs use the same destination field. An import without a
+`destination` inherits its declaring document's effective destination; an
+explicit sequence of key strings appends a relative suffix. Nested imports
+append to their immediate parent's effective destination, while filesystem
+paths remain relative to the declaring file. Destinations are application
+mapping paths, independent of filesystem context. The configuration contract
+owns the complete behavior and validation rules.
+
+The small [`Example_inputDestination`](../../examples/consumers/destinations_test.go)
+mounts one mapping source beneath literal collection/prompt keys and loads a
+second input without a destination to show that root placement remains the
+default. Consumers pass paths, not filenames or namespace strings for ConfigKit
+to interpret.
+
+[`LoadCollectionBundle`](../../examples/consumers/destinations.go) assembles an
+embedded collection manifest mounted beneath a literal collection path. The
+entry's YAML places nested metadata imports at the collection root and prompt
+and profile peers beneath their own keys. A higher-priority input changes only
+one prompt's ordinary `version` field. The main document alone supplies the
+application schema version. A prompt's local `extends` relationship resolves
+after all tiers, so `custom` inherits the final `brief` content and fields while
+retaining its own profile selection.
+
+The resolver follows each effective `content_file` origin to its declaring
+source context. This also works for `custom`'s inherited resource. The
+application-owned [`ReadCollectionPromptContent`](../../examples/consumers/destinations.go)
+demonstrates the same lookup after loading; ConfigKit does not open or interpret
+the resource. [`destinations_test.go`](../../examples/consumers/destinations_test.go)
+checks source destinations and authored coordinates, presence, inherited values,
+relative resources, and strict peer conflicts. The
+[configuration contract](../config.md#input-destinations) owns shared placement,
+import, source-location, and ownership behavior.
+
+## Identified prompt sequences
+
+[`LoadPromptCatalog`](../../examples/consumers/prompt_catalog.go) uses one
+definition registration for the literal `prompts` path and `id` field. Its
+load-wide filename defaults accept two patterns, an explicit directory
+declaration overrides those patterns, and an exclusion omits one matching
+candidate. A nested ordinary helper import supplies an item's ID. A separate
+empty optional directory search is read from `SourceInfo.ImportSelections`.
+The executable example prints both the scheduled prompt order and the single
+empty-search outcome; the behavior test also follows a mounted resource through
+its effective metadata origin to the declaring source context. The completed
+main sequence replaces the lower-priority default collection as one value.
+The maintained [executable example and behavior tests](../../examples/consumers/prompt_catalog_test.go)
+show the scheduled result, empty-search inspection, excluded candidate, and
+resource lookup without a consumer-side discovery pass. The corresponding
+[complex fixture](../../examples/consumers/testdata/prompt-catalog/config.yml)
+and [simple required-file fixture](../../examples/consumers/testdata/prompt-catalog/simple.yml)
+are complete local examples.
+
+[`LoadRequiredPromptCatalog`](../../examples/consumers/prompt_catalog.go) and
+[`Example_requiredPromptCatalog`](../../examples/consumers/prompt_catalog_test.go)
+keep the smaller adoption path visible: one explicitly required file import,
+the same identity registration, and the same application validation.
+
+The example's `validatePromptCatalog` remains application-owned. It checks that
+prompt names and message lists satisfy this consumer's requirements after
+ConfigKit has composed, projected, and decoded the configuration. It does not
+add recursive record merging, identity-based overlay, or final uniqueness checks
+after defaults and resolution. See the [strict composition contract](../config.md#strict-document-composition),
+[import syntax](../config.md#explicit-imports), and
+[ADR 0014](../adr/0014-sequence-placement-and-composition.md) for the exact
+boundaries.
+
 Known module catalogs should use canonical object entries with a module identifier
 and typed options. A small union struct is sufficient when the available option
 fields are known; validation selects which fields apply to each module. Runtime-
@@ -100,6 +267,16 @@ The examples preserve domain meaning while adopting one configuration convention
   blocked empty variables remain blocked, and ordinary `Load` never changes the
   process environment.
 
+### Selected-overlay migration
+
+To adopt the current selected-overlay API, rename `Options.Profile`,
+`LayerProfile`, and `CodeProfile*` references to their `Overlay` equivalents; move
+selected files from `profiles/<name>.yml` to `overlays/<name>.yml`; update any
+handling of the `overlay` layer and `overlay.*` diagnostic strings; then rebuild.
+The [configuration contract](../config.md#selected-overlays) is the canonical
+behavioral reference. Application-owned profile concepts, such as `model_profile`,
+are unchanged, and any CLI flag rename remains the application's decision.
+
 These examples are migration evidence, not completed changes to Narratio, Notarius,
 WeatherReporter, or Seriatim. Updating those repositories, their deployed YAML, CLI
 wiring, runtime catalogs, and operational documentation remains separate downstream
@@ -113,9 +290,13 @@ focused behavioral owner:
 | Roadmap surface | Maintained owner |
 | --- | --- |
 | `Load`, `NoSource`, validation, versions, resolution, `SecretEnv`, overrides, and explanation | Executable examples in [`example_test.go`](../../example_test.go) |
-| Source/import/profile pipeline rules and artifact defaults | Narratio-shaped executable example and fixture |
+| Source/import/selected-overlay pipeline rules and artifact defaults | Narratio-shaped executable example and fixture |
 | Canonical module objects and dynamic option handling | WeatherReporter- and Notarius-shaped examples and migration tests |
 | `LoadSecretsDir`, bootstrap reporting, typed lookup, CLI explanation | `Example_weatherStartup` |
+| Native filesystem, grouped underlays, programmatic mapping, retained resource context | `Example_nativeFilesystem`, `Example_layeredCatalog`, and their focused consumer tests |
+| Literal input destinations, root-default compatibility, and mounted peer collections | `Example_inputDestination`, `LoadCollectionBundle`, and `destinations_test.go` |
+| Named sibling inheritance, presence, and derivation | `Example_inheritedPresets` and `TestPresetCatalogOneLoadAndDerivation` |
+| Effective configuration diffs and consumer-owned presentation | `ExampleDiff`, `ExampleCompareFiles`, and `ExampleCompareRolloutOverlay` |
 | Results, metadata, diagnostics, effective output, and writer failures | Focused root-package load, provenance, rendering, and explanation tests |
 
 The package-level import-boundary test covers the consumer examples too: supported

docs/development.md

@@ -8,29 +8,61 @@ public behavior.
 ## Initial orientation
 
 1. Inspect the working tree and preserve unrelated changes.
-2. Read the architecture policy for code or design work.
-3. Read the policy, contract, and internal documents applicable to the task.
+2. Use the Policy Reading Guide below to select applicable policies.
+3. Read the contract and internal documents applicable to the task.
 4. Inspect the relevant implementation and behavioral tests before changing them.
 
+## Policy Reading Guide
+
+| Policy | Read before… |
+| --- | --- |
+| [Policy set provenance](policy/meta.md) | Adopting policies, checking upstream revisions, or updating the adopted policy set. |
+| [Agents](policy/agents.md) | Recommending, planning, or implementing repository changes. |
+| [Architecture](policy/architecture.md) | Planning or changing Go library code or its structure. |
+| [Testing](policy/testing.md) | Selecting validation for code changes, or adding, changing, removing, or evaluating tests. |
+| [Documentation](policy/documentation.md) | Writing, revising, or reorganizing documentation. |
+| [OSS](policy/oss.md) | Adding or updating dependencies or incorporated third-party material, or changing licensing or attribution. |
+| [Release](policy/release.md) | Preparing or publishing a release, changing release procedures, or making versioning and compatibility decisions. |
+
+- Read applicable policies before making decisions they govern; multiple policies
+  may apply.
+- Select policies using the task triggers above. Follow cross-references to
+  relevant sections when needed to resolve a concrete question.
+- Reuse policies already available in context; reread when changed or no longer
+  available.
+- Expand reading when the task's scope expands.
+- When revising a policy itself, read that policy and any provisions elsewhere
+  affected by the revision.
+
 ## Task-specific reading guide
 
 | Task | Read |
 | --- | --- |
-| Repository orientation or component responsibility | [Architecture policy](policy/architecture.md) and [component overview](internal/overview.md) |
-| Documentation changes | [Documentation policy](policy/documentation.md) and the owning document |
-| Architecture decisions or future work | [Documentation policy](policy/documentation.md), [deferred scope](roadmap/deferred.md), and relevant ADRs listed below |
-| Code and test changes | [Architecture policy](policy/architecture.md), [testing policy](policy/testing.md), and the owning contract and tests from the map below |
+| Repository orientation or component responsibility | [Component overview](internal/overview.md) |
+| Input destinations | [Configuration contract](config.md#input-destinations), [ADR 0011](adr/0011-input-destinations.md), and the destination tests below |
+| Shared source-version checks | [Configuration contract](config.md#shared-source-version-checks), [ADR 0013](adr/0013-shared-source-version-checks.md), and the version/source-version tests below |
+| Documentation changes | The owning document |
+| Licensing, dependencies, attribution, or release-document changes | [Project grant](../COPYRIGHT) and [attribution](oss/attribution.md); release changes also follow the release procedure |
+| Architecture decisions or future work | [Future features](roadmap/future.md) and relevant ADRs listed below |
+| Code and test changes | The owning contract and tests from the map below |
 | Examples or copyable assets | The owning implemented contract and [consumer adoption guide](consumers/README.md) |
-| Release preparation | [Release policy](policy/release.md) and [release procedure](release.md) |
+| Release preparation | [Release procedure](release.md) |
 
 Accepted decisions:
 
 - [Separate composition and overlay](adr/0001-separate-composition-and-overlay.md)
+- [Check source versions with one shared policy](adr/0013-shared-source-version-checks.md)
 - [Own YAML semantics behind a private parser boundary](adr/0002-own-yaml-semantics.md)
 - [Preserve presence and record application resolution](adr/0003-preserve-presence-and-resolution-history.md)
 - [Secret references and explicit bootstrap](adr/0004-secret-references-and-explicit-bootstrap.md)
 - [Explicit overrides and safe explanation](adr/0005-explicit-overrides-and-safe-explanation.md)
 - [Keep the public facade root-owned](adr/0006-public-facade.md)
+- [Compare frozen effective configuration snapshots](adr/0010-effective-configuration-diff.md)
+- [Place inputs at literal application mapping paths](adr/0011-input-destinations.md)
+- [Share dependency-ready ordering](adr/0015-shared-dependency-ready-ordering.md)
+- [Separate import selection controls](adr/0016-independent-import-controls.md)
+- [Define identity and merge rules for each collection](adr/0017-collection-sequence-policy.md)
+- [Apply CLI overrides after inheritance and decoding](adr/0018-cli-after-structural-resolution.md)
 
 For cross-cutting changes, follow every applicable row. Completed implementation
 and remediation plans are preserved in Git history at `cb5b435`; use current
@@ -41,7 +73,9 @@ git show cb5b435:docs/roadmap/remediation.md
 ```
 
 Release-candidate, publication, and verification work follows the operational
-[release procedure](release.md).
+[release procedure](release.md). Its module-selected ReleaseKit `publish`
+invocation validates the committed candidate and publishes only the intended
+tag; those release checks remain distinct from the everyday development check.
 
 The completed string-and-integer schema-version implementation and remediation
 records are preserved in Git history at `244104f`. The
@@ -56,9 +90,87 @@ and current facade ownership belongs to the [component overview](internal/overvi
 
 ## Baseline validation
 
-Install Go 1.25.0 or a supported patched/newer Go toolchain. Acquire declared
-dependencies explicitly before entering an offline development or validation
-environment:
+The completed filesystem-source and underlay roadmap, original feature request,
+implementation plan, and resolved remediation record are preserved in Git history
+at `b604aa1` under `docs/roadmap/`. Current behavior belongs to the
+[configuration contract](config.md#layered-loading). Built-in sibling
+inheritance is now part of the [configuration contract](config.md#sibling-inheritance).
+The completed request, roadmap, plan, and resolved review findings are preserved
+in Git history at `83c59ea` under `docs/roadmap/`.
+
+The completed overlay rename plan and release-workflow simplification record are
+preserved in Git history at `73ce6e9` under `docs/roadmap/`. Current overlay
+behavior belongs to the [configuration contract](config.md#selected-overlays),
+and release operations belong to the [release procedure](release.md).
+
+The completed ReleaseKit adoption, startup performance request and experiments,
+benchmark evidence, performance audit, implementation plan, and remediation
+record are preserved in Git history at `8e91a2b` under `docs/roadmap/`. For example:
+
+```sh
+git show 8e91a2b:docs/roadmap/startup-benchmark.md
+```
+
+Remaining optimization candidates belong to
+[future performance experiments](roadmap/future.md#3-performance-experiments).
+
+The completed effective-configuration diff roadmap, implementation plan, and
+resolved review findings are preserved in Git history at `1c2610c` under
+`docs/roadmap/`. The
+[configuration contract](config.md#effective-configuration-diffs) owns current
+comparison and disclosure behavior, and
+[ADR 0010](adr/0010-effective-configuration-diff.md) records the decision.
+
+The completed input-destination, relative-import-destination, and shared
+source-version roadmaps, associated consumer request, implementation plan, and
+resolved review findings are preserved in Git history at `e4270c9` under
+`docs/roadmap/`. For example:
+
+```sh
+git show e4270c9:docs/roadmap/implementation.md
+git show e4270c9:docs/roadmap/remediation.md
+```
+
+Current behavior belongs to the configuration contract's
+[input destinations](config.md#input-destinations),
+[imports](config.md#explicit-imports), and
+[shared source-version checks](config.md#shared-source-version-checks) sections.
+Explicit sequence-item placement and registered strict composition are owned by
+the [strict composition](config.md#strict-document-composition) and
+[imports](config.md#explicit-imports) sections. The
+[maintained prompt catalog](consumers/README.md#identified-prompt-sequences)
+shows the current consumer boundary, and [ADR 0014](adr/0014-sequence-placement-and-composition.md)
+records the decision.
+
+The completed sequence-composition roadmap, implementation plan, and resolved
+remediation record are preserved in Git history at `008d2ba` under
+`docs/roadmap/`. For example:
+
+```sh
+git show 008d2ba:docs/roadmap/sequence-composition.md
+git show 008d2ba:docs/roadmap/implementation.md
+git show 008d2ba:docs/roadmap/remediation.md
+```
+
+The completed import-discovery roadmap, implementation plan, and resolved
+remediation record are preserved in Git history at `5747ec4` under
+`docs/roadmap/`. For example:
+
+```sh
+git show 5747ec4:docs/roadmap/imports.md
+git show 5747ec4:docs/roadmap/implementation.md
+git show 5747ec4:docs/roadmap/remediation.md
+```
+
+The subsequent import-integration audit and completed remediation record are
+preserved at `6903c9e:docs/roadmap/remediation.md`; retrieve them with
+`git show 6903c9e:docs/roadmap/remediation.md`.
+
+Potential enhancements are tracked in [future features](roadmap/future.md).
+
+Install the Go version required by `go.mod` or a compatible newer toolchain.
+Acquire declared dependencies explicitly before entering an offline
+development or validation environment:
 
 ```sh
 go mod download
@@ -88,8 +200,10 @@ Use focused checks while iterating, with validation proportionate to the change
 under the testing policy. Documentation-only work normally needs the local-link
 check below and whitespace review; behavior-bearing examples and commands also
 need execution. Use `go test -race ./...` for ownership or concurrency changes.
-The separate [release checker](release.md#4-validate-the-source-candidate) adds
-two-toolchain, module-hygiene, race, and cross-build gates.
+The [release procedure](release.md#4-publish-the-committed-candidate) performs
+its selected validation through one `publish` invocation. The everyday checker
+remains a development gate and is not a separate release prerequisite. CI never
+gates a release.
 
 ## Task-to-test map
 
@@ -100,15 +214,26 @@ code changes. Use the component overview to find the corresponding implementatio
 | Task | Behavioral test owners | Focused command |
 | --- | --- | --- |
 | YAML syntax, diagnostics, parser limits | [Adapter tests](../internal/yaml/parse_test.go), [source translation](../internal/engine/parse_test.go) | `go test ./internal/yaml` and `go test ./internal/engine -run '^TestParseSource'` |
-| File/reader access, imports, globs, containment | [Sources](../internal/engine/source_test.go), [FIFO](../internal/engine/source_fifo_test.go), [imports](../internal/engine/imports_test.go), [globs](../internal/engine/glob_test.go) | `go test ./internal/engine -run 'Source\|Import\|Glob\|Filesystem\|Nonregular\|RegularFile'` |
-| Composition, profiles, CLI precedence | [Composition](../internal/engine/compose_test.go), [overlay](../internal/engine/overlay_test.go), [profiles](../internal/engine/profile_test.go), [overrides](../internal/engine/override_test.go) | `go test ./internal/engine -run 'Compose\|Overlay\|Profile\|Override'` |
+| File/reader access, imports, globs, containment | [Sources](../internal/engine/source_test.go), [FIFO](../internal/engine/source_fifo_test.go), [imports](../internal/engine/imports_test.go), [relative destinations](../internal/engine/destination_import_test.go), [globs](../internal/engine/glob_test.go) | `go test ./internal/engine -run 'Source\|Import\|Glob\|Filesystem\|Nonregular\|RegularFile'` |
+| Import defaults, empty-selection evidence, and dependency-ready ordering | [Defaults and facade evidence](../facade_import_defaults_test.go), [engine import evidence](../internal/engine/import_evidence_test.go), [scheduler](../internal/engine/dependency_order_test.go), [scheduling integrations](../internal/engine/imports_test.go) | `go test ./internal/engine -run 'Import|ScheduleDependencies|Inheritance'`, `go test . -run 'ImportDefaults|FacadeImport'`, and `go test ./examples/consumers` |
+| Native filesystem and mapping sources | [Native acquisition](../internal/engine/native_filesystem_test.go), [filesystem facade](../facade_filesystem_test.go), [mapping facade](../facade_mapping_test.go) | `go test ./internal/engine -run 'NativeFilesystem'` and `go test . -run 'TestFacade'` |
+| Composition, selected overlays, CLI precedence | [Composition](../internal/engine/compose_test.go), [overlay](../internal/engine/overlay_test.go), [selected overlays](../internal/engine/selected_overlay_test.go), [overrides](../internal/engine/override_test.go) | `go test ./internal/engine -run 'Compose\|Overlay\|Override'` |
+| Sibling inheritance, provenance, history, sensitivity | [Expansion](../internal/engine/inheritance_test.go), [destination scopes](../internal/engine/inheritance_walk_test.go), [origins](../internal/engine/inheritance_origin_test.go), [history](../internal/engine/inheritance_history_test.go), [secrets](../internal/engine/inheritance_secret_test.go), [public loading](../facade_inheritance_test.go) | `go test ./internal/engine -run 'Inheritance\|Inherited'` and `go test . -run '^TestFacadeInheritance\|^TestFacadeInheritedSecrets'` |
+| Layered loading and source equivalence | [Grouped lifecycle](../internal/engine/load_request_test.go), [layered facade](../facade_layered_test.go), [equivalence](../facade_equivalence_test.go) | `go test ./internal/engine -run '^TestGrouped'`, `go test . -run '^TestFacadeLayered'`, and `go test . -run '^TestEquivalentSourceForms'` |
+| Input and relative-import destinations, mounted sources, and source ownership | [Destination engine tests](../internal/engine/destination_test.go), [tier tests](../internal/engine/destination_tiers_test.go), [import traversal and version tests](../internal/engine/destination_import_test.go), [accounting tests](../internal/engine/destination_accounting_test.go), [lifecycle inspection tests](../internal/engine/destination_import_lifecycle_test.go), [public source/error tests](../facade_destination_test.go), [resource contexts](../facade_resource_context_test.go), and [consumer examples](../examples/consumers/destinations_test.go) | `go test ./internal/engine -run 'Destination|Import|Glob|Mounted'`, `go test . -run 'Destination|Import|SourceContext|FacadeMounted'`, and `go test ./examples/consumers` |
+| Shared source-version modes and evidence | [Engine source-version tests](../internal/engine/source_version_test.go), [public facade tests](../facade_source_version_mode_test.go), and [consumer example tests](../examples/consumers/layered_test.go) | `go test ./internal/engine -run 'SourceVersion'`, `go test . -run 'SourceVersion'`, and `go test ./examples/consumers` |
 | Schema, defaults, projection, cloning | [Schema](../internal/engine/schema_test.go), [decoding](../internal/engine/decode_test.go), [projection](../internal/engine/projection_test.go) | `go test ./internal/engine -run 'Schema\|Decode\|Projection\|Clone\|Inline\|CompareDocument'` |
 | Lifecycle, resolution, validation, callback ownership | [Loading](../internal/engine/load_test.go), [resolution](../internal/engine/resolution_test.go), [validation](../internal/engine/validation_test.go), [ownership](../internal/engine/callback_ownership_test.go), [independence](../internal/engine/independence_test.go) | `go test ./internal/engine -run 'Load\|Resolution\|Validation\|Validator\|Hook\|Callback\|Parallel'` |
 | Paths, metadata, provenance | [Paths](../internal/engine/path_test.go), [metadata](../internal/engine/metadata_test.go), [provenance](../internal/engine/provenance_test.go) | `go test ./internal/engine -run 'Path\|Metadata\|Provenance\|Queries'` |
+| Passive resource contexts | [Engine contexts](../internal/engine/resource_context_test.go), [public contexts](../facade_resource_context_test.go) | `go test ./internal/engine -run 'Resource'` and `go test . -run 'SourceContext'` |
 | Secrets, bootstrap, redaction, explanation | [References](../internal/engine/secret_reference_test.go), [bootstrap](../internal/engine/secrets_test.go), [rendering](../internal/engine/render_test.go), [explanation](../internal/engine/explain_test.go) | `go test ./internal/engine -run 'Secret\|Bootstrap\|Effective\|Explain\|Explanation\|History\|Sensitivity'` |
-| Consumer examples and migration fixtures | [Consumer tests](../examples/consumers/consumers_test.go), [API examples](../example_test.go) | `go test ./examples/consumers` and `go test . -run '^Example'` |
+| Effective configuration diff | [Engine comparison and masking tests](../internal/engine/diff_test.go), [engine operation tests](../internal/engine/diff_operation_test.go), [root facade tests](../facade_diff_test.go), and [consumer examples](../examples/consumers/diff_test.go) | `go test ./internal/engine -run 'Diff|Explain|History|Secret|Independent'`, `go test . -run 'Diff|FacadeInspection|PublicFacadeBoundary|DependencyImportBoundary'`, `go test ./examples/consumers`, and `go test . -run '^ExampleDiff$'` |
+| Consumer examples and migration fixtures | [Consumer tests](../examples/consumers/consumers_test.go), [native, layered, and ordinary relative-import examples](../examples/consumers/layered_test.go), [mounted collection examples](../examples/consumers/destinations_test.go), [API examples](../example_test.go) | `go test ./examples/consumers` and `go test . -run '^Example'` |
+| Assembly and layered scaling evidence | [Assembly benchmarks](../internal/engine/assembly_bench_test.go), [layered benchmarks](../internal/engine/underlay_bench_test.go) | `go test ./internal/engine -run '^$' -bench '^BenchmarkUnderlay' -benchmem`; repeat with `^BenchmarkCompose` and `^BenchmarkOverlay` |
+| Inheritance scaling evidence | [Expansion benchmarks](../internal/engine/inheritance_bench_test.go) | `go test ./internal/engine -run '^$' -bench '^BenchmarkInheritance' -benchmem` |
 | Documentation links and dependency isolation | [Link checks](../documentation_test.go), [import boundary](../import_boundary_test.go), [public boundary](../public_boundary_test.go) | `go test . -run 'TestLocalDocumentationLinks\|TestDependencyImportBoundary\|TestPublicFacadeBoundary'` |
-| Check-command behavior | [Shell regression](../release_check_test.go) | `go test . -run '^TestCheckScripts'` |
+| Development check behavior | [Development checker test](../development_check_test.go) | `go test . -run '^TestDevelopmentCheckHandlesFilenamesAndStopsOnFailure$'` |
+| Release profile wiring | [Release profile test](../releasekit_profile_test.go) | `go test . -run '^TestReleaseKitProfile$'` |
 
 ## Fuzz validation
 

docs/internal/extraction.md

@@ -137,9 +137,11 @@ permission failures, explicit `conf.d` use, and entry/candidate budgets. This wo
 has no Narratio production equivalent; it uses the selected standard-library
 `path.Match` and bounded `File.ReadDir` adaptation rather than its flat loader.
 
-### Profiles and overlay
+### Upstream profiles and ConfigKit selected overlays
 
-Assigned to selected-profile overlay:
+The following upstream symbols and case names retain their original profile
+terminology. ConfigKit adapts their selected-overlay behavior without renaming the
+upstream evidence:
 
 - Production: `mergeOverlayComposition`, `mergeOverlayNodes`,
   `selectPipelineProfile`, `loadPipelineProfileOverlays`, and
@@ -153,19 +155,21 @@ Assigned to selected-profile overlay:
   `TestLoadPipelineProfilePairResolvesIndependentEffectivePipelines`.
 - Preserve: recursive mapping overlay, atomic scalar/sequence replacement,
   deterministic selection failures, source ownership, and independent results.
-- Adapt: load only the selected fixed profile path; lowercase `default` means no
-  overlay. ConfigKit permits null and cross-kind replacement and does not support
-  declared defaults, inheritance, arbitrary profile paths, or validating every
+- Adapt: load only the selected fixed overlay path; lowercase `default` means no
+  selection. ConfigKit permits null and cross-kind replacement and does not support
+  declared defaults, inheritance, arbitrary overlay paths, or validating every
   unselected overlay.
 
-Implementation record (2026-09-06): `overlayDocuments`, `mergeOverlayMappings`,
-`loadProfileDocument`, and `profileDirectiveDiagnostics` adapt the assigned merge,
-selection, loading, and ownership mechanics. `TestOverlayDocumentsRecursesMappingsAndReplacesAtomicSubtrees`,
-`TestLoadAppliesSelectedProfileAfterBaseAndBeforeDefaults`, profile selection/error,
-directive/version, containment, and budget cases adapt the assigned tests. ConfigKit
+Implementation record (2026-09-06, updated 2026-09-23): `overlayDocuments`,
+`mergeOverlayMappings`, `loadSelectedOverlayDocument`, and
+`overlayDirectiveDiagnostics` adapt the assigned merge, selection, loading, and
+ownership mechanics. `TestOverlayDocumentsRecursesMappingsAndReplacesAtomicSubtrees`,
+`TestLoadAppliesSelectedOverlayAfterBaseAndBeforeDefaults`, selected-overlay
+selection/error, directive/version, containment, and budget cases adapt the assigned
+tests. ConfigKit
 uses exactly one fixed selected `.yml` source, permits null and cross-kind replacement,
-retains displaced descendant history, rejects any profile directive or version-path
-contribution, and counts the profile in aggregate budgets. Narratio's declared
+retains displaced descendant history, rejects any overlay directive or version-path
+contribution, and counts the overlay in aggregate budgets. Narratio's declared
 default, multi-overlay validation, inheritance-like application conventions, and
 null/kind restrictions were not retained.
 
@@ -196,7 +200,7 @@ previously extracted source claims, overlay displacement records, typed baseline
 and audited resolution snapshots in `load_metadata.go`. The representative
 `TestLayeredProvenanceAgreesWithResolvedSnapshot` adapts Narratio's stable ownership,
 profile ownership, deterministic effective projection, and side-effect-free
-inspection cases across one default/import/profile/resolution workflow. ConfigKit
+inspection cases across one default/import/overlay/resolution workflow. ConfigKit
 adds chronological descendant history for parent and whole-sequence operations,
 composite empty-mapping origins, canonical resolution reasons/dependencies, private
 snapshot agreement, defensive nested copies, and concurrent load isolation. Value

docs/internal/imports.md

@@ -0,0 +1,80 @@
+# Import Discovery and Scheduling Internals
+
+The [configuration contract](../config.md#explicit-imports) owns selector syntax,
+defaults, requiredness, placement, inspection evidence, and observable ordering.
+This note records implementation flow and private identities for maintainers.
+
+## Discovery and acquisition
+
+Each source's root directive is parsed into presence-aware declarations and
+resolved against load-wide defaults. Candidate selection preserves a normalized
+namespace path separately from canonical access identity: overlapping roots
+coalesce the same candidate spelling in that namespace, while aliases reached
+through distinct spellings retain separate inclusion routes for admission.
+Directory searches walk each distinct normalized root once, apply all filename
+patterns to discovered names, and form a set union of matching candidates across
+overlapping roots. Exclusions are applied as set difference after discovery and
+candidate normalization. They do not suppress discovery or resource errors.
+
+Directory traversal tracks canonical directories on the active traversal path.
+It adds an identity when descending and removes it when the frame is popped, so
+only a back-edge to an active directory is a traversal cycle; a later alias is
+not mistaken for one. This is separate from import dependency-cycle detection.
+For host sources, discovery uses rooted `Lstat` and canonical containment. Native
+sources use `fs.ReadLinkFS` inspection when supplied; otherwise traversal follows
+the capabilities and behavior of the provided `fs.FS`, without claiming a
+general sandbox guarantee.
+
+Acquisition remains authored: declarations are visited in source order and each
+selected candidate set in namespace-path lexical order. This preserves source
+records, IDs, source-local version checks, and callback ordering. The acquired
+inclusion graph is retained separately from composition scopes so cycles that
+cross item boundaries remain detectable. Source discovery does not itself choose
+sequence assembly order.
+
+## Stable keys and scheduling
+
+The shared scheduler compares immutable tuples component by component using Go's
+bytewise string order; if one tuple is a prefix of another, the shorter tuple
+sorts first. The import key is
+`(namespace-kind, namespace-anchor, resource-kind, resource-path)`:
+
+- Namespace kind is `host` or `native`. Host resource paths are canonical
+  absolute paths rendered with slash separators; native paths retain their
+  namespace-relative identity.
+- Anchors come from explicit input positions, not discovery order: main is
+  `(0,0)` and underlay/input positions are `(group+1,input+1)`, each encoded as
+  fixed-width decimal. Host accesses in one scheduled tier share its host anchor.
+  A native filesystem namespace uses its earliest explicit selecting position
+  in that tier. Filesystem tokens distinguish namespace equality, not sort order,
+  so separately constructed filesystems with the same label remain distinct.
+- Resource kind is `file` or `synthetic`. A synthetic root uses its explicit
+  input anchor and an empty resource path rather than a display label or token
+  address.
+
+Inheritance keys are
+`("definition", scope-pointer, definition-pointer)`, where both pointers use
+canonical escaped semantic path spelling. Source IDs are not part of the key.
+
+Both integrations use the same Kahn scheduler: it deduplicates dependency edges,
+keeps currently ready nodes in a min-heap, and selects the smallest ready key
+after every completion. Its scheduling cost is `O(E + V log V)` time and
+`O(V + E)` additional memory, excluding key construction, tuple copying, and
+tuple comparisons. On a stalled graph, deterministic DFS over the residual graph
+returns a closed cycle rotated to its smallest key; blocked dependents are not
+reported as cycle members.
+
+Import composition schedules a main tier or one underlay group as a single
+strict-scope graph. Item contributions own nested local strict-mapping scopes;
+their helpers complete before the item is placed in its outer scope. Inheritance
+uses the same ready-node rule inside each sibling mapping scope, while preserving
+the authored mapping order in its output. These scopes do not flatten into one
+global graph, and independent precedence tiers remain separate.
+
+The main owners are [`import_selection.go`](../../internal/engine/import_selection.go),
+[`directory_walk.go`](../../internal/engine/directory_walk.go),
+[`import_graph.go`](../../internal/engine/import_graph.go),
+[`dependency_order.go`](../../internal/engine/dependency_order.go), and
+[`inheritance_schedule.go`](../../internal/engine/inheritance_schedule.go).
+Their focused behavior is covered by import-selection, scheduler, and scheduling
+integration tests under `internal/engine`.

docs/internal/overview.md

@@ -7,23 +7,43 @@ private state; public signatures and fields do not expose engine types.
 
 Implemented component owners:
 
-- `internal/engine` owns source acquisition, composition, schema inspection,
-  decoding, lifecycle coordination, callback auditing, provenance/history,
-  secret resolution and bootstrap, and safe inspection rendering. It never
-  imports the root package.
-- Root `Source`, `Path`, `Secret`, records, lifecycle callbacks, results, and
-  bootstrap reports are root-defined wrappers or records. Callback wrappers pass
-  only live engine context during a callback; `Result.Config` is detached and its
-  private inspection remains frozen independently.
+- `internal/engine` owns host and native-filesystem source acquisition,
+  grouped strict composition, recursive precedence assembly, per-input mapping
+  and sequence-item placement, definition-registered distinct-ID sequence
+  composition within strict tiers, and immediate-parent-relative import
+  placement within that assembly. Higher-priority sequences replace lower
+  sequences as whole values. The engine also owns dependency-ready,
+  destination-local sibling inheritance before typed decoding, plus shared
+  source-version checking with sanitized declaration evidence.
+  Explicit import discovery supports declared paths, directory searches, and
+  globs; load-wide defaults and source-local empty-selection evidence cross the
+  root facade without changing acquisition order. Import and inheritance
+  resolution share dependency-ready scheduling within their composition scopes.
+  It retains each source's effective placement alongside physical source
+  identity and owns schema
+  inspection, lifecycle coordination, callback auditing,
+  provenance/history, passive source-context retention, secret resolution and
+  bootstrap, safe inspection rendering, and effective-value comparison over
+  frozen inspection snapshots. It never imports the root package.
+- Root `Source`, `Filesystem`, `Input`, `Underlay`, `Path`, `Secret`, records,
+  lifecycle callbacks, results, and bootstrap reports are root-defined wrappers
+  or records. Callback wrappers pass only live engine context during a callback;
+  `Result.Config` is detached and its private inspection remains frozen
+  independently.
 - Root conversion files explicitly cross the path, record, callback, secret, and
   error boundaries. There are no public aliases, anonymous engine embeddings, or
   public engine bridge methods.
+- Root `Diff` and `Difference` provide the public comparison boundary and
+  convert detached engine records to root-owned paths and snapshots; consumer
+  examples own presentation of the safe changed regions.
 - `internal/document` owns the parser-independent ordered tree and cloning.
   `internal/yaml` is the sole upstream YAML adapter and parser/dumper boundary.
-- Engine mechanism tests own source, composition, decoding, callback, provenance,
-  secret, history, rendering, and bootstrap behavior. Root tests own public API
-  compatibility, facade integration, documentation links, and dependency/public
-  boundary checks. Consumer examples remain under `examples/consumers`.
+- Engine mechanism tests own source, composition, inheritance, decoding,
+  callback, provenance,
+  source-context, secret, history, rendering, and bootstrap behavior. Root tests
+  own public API compatibility, facade integration, documentation links, and
+  dependency/public boundary checks. Consumer examples remain under
+  `examples/consumers`.
 
 Dependency direction is root to engine to document/YAML support. The root may
 import only `internal/engine` among project internal production packages; engine
@@ -32,8 +52,27 @@ these constraints and rejects exported aliases, fields, or signatures involving
 internal types.
 
 The [configuration contract](../config.md) owns implemented YAML and lifecycle
-semantics. The [consumer adoption guide](../consumers/README.md) owns application
-guidance. The [architecture policy](../policy/architecture.md) owns contributor
-boundaries, and [ADR 0006](../adr/0006-public-facade.md) records the facade
+semantics, including [input destinations](../config.md#input-destinations),
+[shared source-version checks](../config.md#shared-source-version-checks), and
+[effective configuration diffs](../config.md#effective-configuration-diffs).
+The [strict composition contract](../config.md#strict-document-composition)
+owns sequence placement, registration, replacement, and history behavior; the
+[consumer guide](../consumers/README.md#identified-prompt-sequences) shows the
+maintained example.
+The [consumer adoption guide](../consumers/README.md) owns application guidance.
+The [architecture policy](../policy/architecture.md) owns contributor
+boundaries; [ADR 0006](../adr/0006-public-facade.md) records the facade decision
+and [ADR 0010](../adr/0010-effective-configuration-diff.md) records the diff
+decision; [ADR 0011](../adr/0011-input-destinations.md) records the placement
 decision. See the [development guide](../development.md) for validation commands
-and [release procedure](../release.md) for source-candidate gates.
+and [release procedure](../release.md) for source-candidate gates. [ADR
+0009](../adr/0009-sibling-inheritance.md) records why inheritance is a built-in
+post-assembly structural operation. [ADR 0012](../adr/0012-relative-import-destinations.md)
+records the relative-import extension; [ADR 0013](../adr/0013-shared-source-version-checks.md)
+records the source-version decision.
+[ADR 0014](../adr/0014-sequence-placement-and-composition.md) records explicit
+sequence placement and composition. [ADR 0015](../adr/0015-shared-dependency-ready-ordering.md)
+records the shared dependency-ready ordering decision; [ADR
+0016](../adr/0016-independent-import-controls.md) records the independent
+selection controls. The [import internals](imports.md) document candidate and
+namespace identities, discovery and scheduling, and the bounded algorithms.

docs/oss/attribution.md

@@ -0,0 +1,43 @@
+# Open Source Attribution Notices
+
+## Standard Repository Policies
+
+Material: `docs/policy/release.md`, adapted from
+`e833faddf120c789a9b45348edaf75720de16348`; `docs/policy/meta.md`, `docs/policy/oss.md`,
+`docs/policy/documentation.md`, `docs/policy/agents.md`,
+`docs/policy/architecture.md`, `docs/policy/testing.md`; root `AGENTS.md`; and the
+Policy Reading Guide in `docs/development.md`, adapted from
+`d84a38ebab26f048cafbe59a2246700c35d29843` for this repository.
+
+Copyright (C) 2026 Eric Rakestraw
+
+SPDX-License-Identifier: CC-BY-SA-4.0
+
+License and notices: [CC BY-SA 4.0](licenses/CC-BY-SA-4.0.txt) |
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/COPYRIGHT) |
+[LICENSE](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/LICENSE) |
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/policy/src/commit/e833faddf120c789a9b45348edaf75720de16348/COPYRIGHT) |
+[LICENSE](https://gitea.maximumdirect.net/eric/policy/src/commit/e833faddf120c789a9b45348edaf75720de16348/LICENSE).
+
+[Source](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843).
+
+[Source](https://gitea.maximumdirect.net/eric/policy/src/commit/e833faddf120c789a9b45348edaf75720de16348).
+
+## go-yaml v4
+
+Selected module: `go.yaml.in/yaml/v4` at `v4.0.0-rc.6`.
+
+Copyright 2011-2019 Canonical Ltd
+
+Copyright 2025 The go-yaml Project Contributors
+
+Copyright 2006-2010 Kirill Simonov
+
+SPDX-License-Identifier: Apache-2.0 AND MIT
+
+License and notices: [Apache 2.0](licenses/Apache-2.0.txt) |
+[MIT](licenses/MIT.txt) |
+[NOTICE](https://github.com/yaml/go-yaml/blob/1c17b9cee72bf81ae73a8be50fbd7dd3a9985125/NOTICE) |
+[LICENSE](https://github.com/yaml/go-yaml/blob/1c17b9cee72bf81ae73a8be50fbd7dd3a9985125/LICENSE).
+
+[Source](https://github.com/yaml/go-yaml/tree/1c17b9cee72bf81ae73a8be50fbd7dd3a9985125).

docs/oss/licenses/Apache-2.0.txt

@@ -0,0 +1,202 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

docs/oss/licenses/CC-BY-SA-4.0.txt

@@ -0,0 +1,170 @@
+Creative Commons Attribution-ShareAlike 4.0 International
+
+ Creative Commons Corporation (“Creative Commons”) is not a law firm and does not provide legal services or legal advice. Distribution of Creative Commons public licenses does not create a lawyer-client or other relationship. Creative Commons makes its licenses and related information available on an “as-is” basis. Creative Commons gives no warranties regarding its licenses, any material licensed under their terms and conditions, or any related information. Creative Commons disclaims all liability for damages resulting from their use to the fullest extent possible.
+
+Using Creative Commons Public Licenses
+
+Creative Commons public licenses provide a standard set of terms and conditions that creators and other rights holders may use to share original works of authorship and other material subject to copyright and certain other rights specified in the public license below. The following considerations are for informational purposes only, are not exhaustive, and do not form part of our licenses.
+
+Considerations for licensors: Our public licenses are intended for use by those authorized to give the public permission to use material in ways otherwise restricted by copyright and certain other rights. Our licenses are irrevocable. Licensors should read and understand the terms and conditions of the license they choose before applying it. Licensors should also secure all rights necessary before applying our licenses so that the public can reuse the material as expected. Licensors should clearly mark any material not subject to the license. This includes other CC-licensed material, or material used under an exception or limitation to copyright. More considerations for licensors.
+
+Considerations for the public: By using one of our public licenses, a licensor grants the public permission to use the licensed material under specified terms and conditions. If the licensor’s permission is not necessary for any reason–for example, because of any applicable exception or limitation to copyright–then that use is not regulated by the license. Our licenses grant only permissions under copyright and certain other rights that a licensor has authority to grant. Use of the licensed material may still be restricted for other reasons, including because others have copyright or other rights in the material. A licensor may make special requests, such as asking that all changes be marked or described.
+
+Although not required by our licenses, you are encouraged to respect those requests where reasonable. More considerations for the public.
+
+Creative Commons Attribution-ShareAlike 4.0 International Public License
+
+By exercising the Licensed Rights (defined below), You accept and agree to be bound by the terms and conditions of this Creative Commons Attribution-ShareAlike 4.0 International Public License ("Public License"). To the extent this Public License may be interpreted as a contract, You are granted the Licensed Rights in consideration of Your acceptance of these terms and conditions, and the Licensor grants You such rights in consideration of benefits the Licensor receives from making the Licensed Material available under these terms and conditions.
+
+Section 1 – Definitions.
+
+     a.	Adapted Material means material subject to Copyright and Similar Rights that is derived from or based upon the Licensed Material and in which the Licensed Material is translated, altered, arranged, transformed, or otherwise modified in a manner requiring permission under the Copyright and Similar Rights held by the Licensor. For purposes of this Public License, where the Licensed Material is a musical work, performance, or sound recording, Adapted Material is always produced where the Licensed Material is synched in timed relation with a moving image.
+
+     b.	Adapter's License means the license You apply to Your Copyright and Similar Rights in Your contributions to Adapted Material in accordance with the terms and conditions of this Public License.
+
+     c.	BY-SA Compatible License means a license listed at creativecommons.org/compatiblelicenses, approved by Creative Commons as essentially the equivalent of this Public License.
+
+     d.	Copyright and Similar Rights means copyright and/or similar rights closely related to copyright including, without limitation, performance, broadcast, sound recording, and Sui Generis Database Rights, without regard to how the rights are labeled or categorized. For purposes of this Public License, the rights specified in Section 2(b)(1)-(2) are not Copyright and Similar Rights.
+
+     e.	Effective Technological Measures means those measures that, in the absence of proper authority, may not be circumvented under laws fulfilling obligations under Article 11 of the WIPO Copyright Treaty adopted on December 20, 1996, and/or similar international agreements.
+
+     f.	Exceptions and Limitations means fair use, fair dealing, and/or any other exception or limitation to Copyright and Similar Rights that applies to Your use of the Licensed Material.
+
+     g.	License Elements means the license attributes listed in the name of a Creative Commons Public License. The License Elements of this Public License are Attribution and ShareAlike.
+
+     h.	Licensed Material means the artistic or literary work, database, or other material to which the Licensor applied this Public License.
+
+     i.	Licensed Rights means the rights granted to You subject to the terms and conditions of this Public License, which are limited to all Copyright and Similar Rights that apply to Your use of the Licensed Material and that the Licensor has authority to license.
+
+     j.	Licensor means the individual(s) or entity(ies) granting rights under this Public License.
+
+     k.	Share means to provide material to the public by any means or process that requires permission under the Licensed Rights, such as reproduction, public display, public performance, distribution, dissemination, communication, or importation, and to make material available to the public including in ways that members of the public may access the material from a place and at a time individually chosen by them.
+
+     l.	Sui Generis Database Rights means rights other than copyright resulting from Directive 96/9/EC of the European Parliament and of the Council of 11 March 1996 on the legal protection of databases, as amended and/or succeeded, as well as other essentially equivalent rights anywhere in the world.
+
+     m.	You means the individual or entity exercising the Licensed Rights under this Public License. Your has a corresponding meaning.
+
+Section 2 – Scope.
+
+     a.	License grant.
+
+          1. Subject to the terms and conditions of this Public License, the Licensor hereby grants You a worldwide, royalty-free, non-sublicensable, non-exclusive, irrevocable license to exercise the Licensed Rights in the Licensed Material to:
+
+               A. reproduce and Share the Licensed Material, in whole or in part; and
+
+               B. produce, reproduce, and Share Adapted Material.
+
+          2. Exceptions and Limitations. For the avoidance of doubt, where Exceptions and Limitations apply to Your use, this Public License does not apply, and You do not need to comply with its terms and conditions.
+
+          3. Term. The term of this Public License is specified in Section 6(a).
+
+          4. Media and formats; technical modifications allowed. The Licensor authorizes You to exercise the Licensed Rights in all media and formats whether now known or hereafter created, and to make technical modifications necessary to do so. The Licensor waives and/or agrees not to assert any right or authority to forbid You from making technical modifications necessary to exercise the Licensed Rights, including technical modifications necessary to circumvent Effective Technological Measures. For purposes of this Public License, simply making modifications authorized by this Section 2(a)(4) never produces Adapted Material.
+
+          5. Downstream recipients.
+
+               A. Offer from the Licensor – Licensed Material. Every recipient of the Licensed Material automatically receives an offer from the Licensor to exercise the Licensed Rights under the terms and conditions of this Public License.
+
+               B. Additional offer from the Licensor – Adapted Material. Every recipient of Adapted Material from You automatically receives an offer from the Licensor to exercise the Licensed Rights in the Adapted Material under the conditions of the Adapter’s License You apply.
+
+               C. No downstream restrictions. You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, the Licensed Material if doing so restricts exercise of the Licensed Rights by any recipient of the Licensed Material.
+
+          6. No endorsement. Nothing in this Public License constitutes or may be construed as permission to assert or imply that You are, or that Your use of the Licensed Material is, connected with, or sponsored, endorsed, or granted official status by, the Licensor or others designated to receive attribution as provided in Section 3(a)(1)(A)(i).
+
+     b.	Other rights.
+
+          1. Moral rights, such as the right of integrity, are not licensed under this Public License, nor are publicity, privacy, and/or other similar personality rights; however, to the extent possible, the Licensor waives and/or agrees not to assert any such rights held by the Licensor to the limited extent necessary to allow You to exercise the Licensed Rights, but not otherwise.
+
+          2. Patent and trademark rights are not licensed under this Public License.
+
+          3. To the extent possible, the Licensor waives any right to collect royalties from You for the exercise of the Licensed Rights, whether directly or through a collecting society under any voluntary or waivable statutory or compulsory licensing scheme. In all other cases the Licensor expressly reserves any right to collect such royalties.
+
+Section 3 – License Conditions.
+
+Your exercise of the Licensed Rights is expressly made subject to the following conditions.
+
+     a.	Attribution.
+
+          1. If You Share the Licensed Material (including in modified form), You must:
+
+               A. retain the following if it is supplied by the Licensor with the Licensed Material:
+
+                    i.	identification of the creator(s) of the Licensed Material and any others designated to receive attribution, in any reasonable manner requested by the Licensor (including by pseudonym if designated);
+
+                    ii.	a copyright notice;
+
+                    iii. a notice that refers to this Public License;
+
+                    iv.	a notice that refers to the disclaimer of warranties;
+
+                    v.	a URI or hyperlink to the Licensed Material to the extent reasonably practicable;
+
+               B. indicate if You modified the Licensed Material and retain an indication of any previous modifications; and
+
+               C. indicate the Licensed Material is licensed under this Public License, and include the text of, or the URI or hyperlink to, this Public License.
+
+          2. You may satisfy the conditions in Section 3(a)(1) in any reasonable manner based on the medium, means, and context in which You Share the Licensed Material. For example, it may be reasonable to satisfy the conditions by providing a URI or hyperlink to a resource that includes the required information.
+
+          3. If requested by the Licensor, You must remove any of the information required by Section 3(a)(1)(A) to the extent reasonably practicable.
+
+     b.	ShareAlike.In addition to the conditions in Section 3(a), if You Share Adapted Material You produce, the following conditions also apply.
+
+          1. The Adapter’s License You apply must be a Creative Commons license with the same License Elements, this version or later, or a BY-SA Compatible License.
+
+          2. You must include the text of, or the URI or hyperlink to, the Adapter's License You apply. You may satisfy this condition in any reasonable manner based on the medium, means, and context in which You Share Adapted Material.
+
+          3. You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, Adapted Material that restrict exercise of the rights granted under the Adapter's License You apply.
+
+Section 4 – Sui Generis Database Rights.
+
+Where the Licensed Rights include Sui Generis Database Rights that apply to Your use of the Licensed Material:
+
+     a.	for the avoidance of doubt, Section 2(a)(1) grants You the right to extract, reuse, reproduce, and Share all or a substantial portion of the contents of the database;
+
+     b.	if You include all or a substantial portion of the database contents in a database in which You have Sui Generis Database Rights, then the database in which You have Sui Generis Database Rights (but not its individual contents) is Adapted Material, including for purposes of Section 3(b); and
+
+     c.	You must comply with the conditions in Section 3(a) if You Share all or a substantial portion of the contents of the database.
+For the avoidance of doubt, this Section 4 supplements and does not replace Your obligations under this Public License where the Licensed Rights include other Copyright and Similar Rights.
+
+Section 5 – Disclaimer of Warranties and Limitation of Liability.
+
+     a.	Unless otherwise separately undertaken by the Licensor, to the extent possible, the Licensor offers the Licensed Material as-is and as-available, and makes no representations or warranties of any kind concerning the Licensed Material, whether express, implied, statutory, or other. This includes, without limitation, warranties of title, merchantability, fitness for a particular purpose, non-infringement, absence of latent or other defects, accuracy, or the presence or absence of errors, whether or not known or discoverable. Where disclaimers of warranties are not allowed in full or in part, this disclaimer may not apply to You.
+
+     b.	To the extent possible, in no event will the Licensor be liable to You on any legal theory (including, without limitation, negligence) or otherwise for any direct, special, indirect, incidental, consequential, punitive, exemplary, or other losses, costs, expenses, or damages arising out of this Public License or use of the Licensed Material, even if the Licensor has been advised of the possibility of such losses, costs, expenses, or damages. Where a limitation of liability is not allowed in full or in part, this limitation may not apply to You.
+
+     c.	The disclaimer of warranties and limitation of liability provided above shall be interpreted in a manner that, to the extent possible, most closely approximates an absolute disclaimer and waiver of all liability.
+
+Section 6 – Term and Termination.
+
+     a.	This Public License applies for the term of the Copyright and Similar Rights licensed here. However, if You fail to comply with this Public License, then Your rights under this Public License terminate automatically.
+
+     b.	Where Your right to use the Licensed Material has terminated under Section 6(a), it reinstates:
+
+          1. automatically as of the date the violation is cured, provided it is cured within 30 days of Your discovery of the violation; or
+
+          2. upon express reinstatement by the Licensor.
+
+     c.	For the avoidance of doubt, this Section 6(b) does not affect any right the Licensor may have to seek remedies for Your violations of this Public License.
+
+     d.	For the avoidance of doubt, the Licensor may also offer the Licensed Material under separate terms or conditions or stop distributing the Licensed Material at any time; however, doing so will not terminate this Public License.
+
+     e.	Sections 1, 5, 6, 7, and 8 survive termination of this Public License.
+
+Section 7 – Other Terms and Conditions.
+
+     a.	The Licensor shall not be bound by any additional or different terms or conditions communicated by You unless expressly agreed.
+
+     b.	Any arrangements, understandings, or agreements regarding the Licensed Material not stated herein are separate from and independent of the terms and conditions of this Public License.
+
+Section 8 – Interpretation.
+
+     a.	For the avoidance of doubt, this Public License does not, and shall not be interpreted to, reduce, limit, restrict, or impose conditions on any use of the Licensed Material that could lawfully be made without permission under this Public License.
+
+     b.	To the extent possible, if any provision of this Public License is deemed unenforceable, it shall be automatically reformed to the minimum extent necessary to make it enforceable. If the provision cannot be reformed, it shall be severed from this Public License without affecting the enforceability of the remaining terms and conditions.
+
+     c.	No term or condition of this Public License will be waived and no failure to comply consented to unless expressly agreed to by the Licensor.
+
+     d.	Nothing in this Public License constitutes or may be interpreted as a limitation upon, or waiver of, any privileges and immunities that apply to the Licensor or You, including from the legal processes of any jurisdiction or authority.
+
+Creative Commons is not a party to its public licenses. Notwithstanding, Creative Commons may elect to apply one of its public licenses to material it publishes and in those instances will be considered the “Licensor.” Except for the limited purpose of indicating that material is shared under a Creative Commons public license or as otherwise permitted by the Creative Commons policies published at creativecommons.org/policies, Creative Commons does not authorize the use of the trademark “Creative Commons” or any other trademark or logo of Creative Commons without its prior written consent including, without limitation, in connection with any unauthorized modifications to any of its public licenses or any other arrangements, understandings, or agreements concerning use of licensed material. For the avoidance of doubt, this paragraph does not form part of the public licenses.
+
+Creative Commons may be contacted at creativecommons.org.

docs/oss/licenses/MIT.txt

@@ -0,0 +1,18 @@
+MIT License
+
+Copyright (c) <year> <copyright holders>
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of this software and
+associated documentation files (the "Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the
+following conditions:
+
+The above copyright notice and this permission notice shall be included in all copies or substantial
+portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT
+LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO
+EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
+IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+USE OR OTHER DEALINGS IN THE SOFTWARE.

docs/policy/agents.md

@@ -0,0 +1,91 @@
+# Agent Policy
+
+**Base Policy Version:** 1.3
+
+**Last Updated**: 2026-09-28
+
+**Upstream Source:** [Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/docs/policy/agents.md)
+
+## Purpose
+
+Agents must read and follow these rules.
+
+## Implementation Discipline
+
+**KEY PRINCIPLES:** Write the simplest idiomatic code that satisfies the user's
+request. Do not implicitly strengthen requirements, restrictions, or guarantees
+beyond what the user actually requested.
+
+**Keep implementation machinery proportionate.** Use straightforward,
+idiomatic code and introduce only the machinery needed to support the requested
+behavior and flexibility. Optimize for conceptual simplicity and
+maintainability, not for the fewest lines, fewest options, or narrowest product
+surface.
+
+**Do not invent restrictions.** Do not reject, prohibit, constrain, cap,
+harden, or otherwise narrow behavior unless required by the user, an explicit
+repository contract, or an unavoidable correctness constraint. Preserve
+operator and caller choice where policy has not been established. If an
+additional restriction appears advisable, present it as a decision rather than
+silently implementing it.
+
+A correctness constraint must follow from the requested behavior or an
+established invariant, not from the convenience of the chosen implementation.
+Explain the specific failure it prevents and why a less restrictive
+implementation cannot satisfy the contract.
+
+**Do not invent guarantees.** Do not strengthen public contracts, guarantees,
+or invariants beyond those expressly requested by the user.
+
+**Use the standard library.** For example: in almost every case, the best
+approach to copying files is to use the standard library. Writing custom code to
+implement power-loss protection at the application level, or to guard against
+bad actors with privileged and concurrent access to the host's filesystem,
+is wasteful, unnecessary, and not idiomatic. Stick to the simple, standard, and
+idiomatic approach unless another path is expressly requested by the user.
+
+**An implementation detail is not repository policy.** Tests and documentation
+must not elevate incidental behavior into a durable guarantee or restriction
+unless that behavior was expressly selected as part of the contract. Existing
+explicit contracts remain authoritative; changing or weakening one requires a
+deliberate decision rather than being inferred from this principle.
+
+Agent-authored plans, tests, documentation, and prior implementations are not
+evidence that an additional restriction was deliberately approved. Identify the
+authorizing requirement or decision; when that authority is unclear, surface
+the ambiguity rather than treating repetition as approval.
+
+## Git Commit Policy
+
+By default, commit each completed, nonempty unit of work using a
+concise, plain-English message describing what changed. This policy is standing
+authorization to create those local commits; a separate user request is not
+required. An explicit user instruction to leave work uncommitted takes precedence.
+
+A unit of work is a focused, reviewable outcome, including its relevant tests and
+documentation.
+Complete the appropriate validation before committing. If work is incomplete or
+validation is blocked, report that state accurately.
+
+Before committing:
+
+1. Run `git status --short` and review the relevant diff, including any staged
+   changes. Identify changes that predate the current unit of work.
+2. Exclude secrets, local environment files, generated junk, and unrelated changes.
+3. Stage only the files or hunks belonging to the completed unit. Use `git add .`
+   only when every included change belongs to that unit. Preserve unrelated
+   staged and unstaged work.
+4. Review `git diff --staged` and run appropriate checks for the change. Reuse
+   successful checks for unchanged content.
+5. Review `git log --oneline -n 10` and use
+   `git commit -m "<concise plain-English commit message>"`.
+6. Report the commit identifier and any remaining uncommitted work.
+
+If the unit cannot be safely separated from pre-existing work, leave it uncommitted
+and explain the overlap rather than silently including that work. If a commit
+fails, preserve the changes and report the failure; do not bypass hooks.
+
+Default commit authorization does not authorize pushing, publishing, amending,
+resetting, reverting, or rewriting history. Those actions require a separate user
+instruction. Do not use `git commit -a` or `git commit --no-verify` to bypass the
+staging and validation discipline above.

docs/policy/architecture.md

@@ -1,18 +1,61 @@
 # Architecture Policy
 
-**Base Policy Version:** 1.0
+**Base Policy Version:** 1.3
 
-**Last Updated:** 2026-09-06
+**Last Updated**: 2026-10-22
 
-## Purpose and applicability
+**Upstream Source:** [Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/docs/policy/architecture.md)
+
+## Purpose And Ownership
+
+The adopted policy defines the application's purpose and the architectural
+constraints contributors must follow. It owns system boundaries, responsibility
+and dependency rules, established invariants, and explicit non-goals.
+
+Describe the concrete package inventory and implementation mechanics in internal
+documentation. Link to canonical public contracts for their exact behavior and
+to ADRs for decision rationale.
 
 ConfigKit is an opinionated Go library for consistent YAML configuration loading
-and resolution across consuming applications. This policy governs contributions.
-The [component overview](../internal/overview.md) owns implementation inventory, the
-[configuration contract](../config.md) owns implemented behavior, and the
-[deferred scope](../roadmap/deferred.md) preserves deferred scope.
+and resolution across consuming applications. The [component overview](../internal/overview.md)
+owns implementation inventory, the [configuration contract](../config.md) owns
+implemented behavior, and the [future feature list](../roadmap/future.md) preserves
+candidates for future consideration.
 
-## Success criteria
+## Adopted Constraints And Implemented Structure
+
+Architectural constraints govern current and future development. Record
+implemented structure accurately, proposals and implementation status in the
+roadmap, and significant decisions and rationale in ADRs. Describe a target
+structure as implemented when its changes land.
+
+Constraints must reflect deliberately selected requirements and decisions.
+Do not promote incidental implementation choices into architectural invariants.
+
+The collection-policy and CLI-ordering requirements below reflect accepted
+[ADR 0017](../adr/0017-collection-sequence-policy.md) and
+[ADR 0018](../adr/0018-cli-after-structural-resolution.md). The
+[sequence roadmap](../roadmap/sequences.md) tracks the work to implement them;
+the configuration contract continues to describe implemented behavior.
+
+## Success Criteria
+
+ConfigKit prioritizes a coherent configuration model: consistency and
+predictability take precedence over flexibility for accommodating individual
+downstream conventions. Consumers should learn one set of rules, not different
+combination or presence semantics depending on the source format or consuming
+application. Distinct operations such as strict composition and intentional
+overlay express distinct purposes within that model.
+
+Consumers should be able to load once and receive a complete, validated
+configuration with inspectable provenance for its final values, including
+defaults and recorded application-derived values. Configuration and metadata
+together must distinguish explicit zero, false, empty, and null input from
+omission; ordinary Go values alone need not encode those distinctions.
+Application-owned resolution belongs within the recorded loading lifecycle, so
+consumers need no subsequent generic merging, defaulting, or provenance
+reconstruction. This does not transfer domain execution or runtime resource
+management into ConfigKit.
 
 1. **Reduce application infrastructure:** Can applications delete substantial
    configuration infrastructure while retaining clear ownership of their domain
@@ -21,8 +64,11 @@ The [component overview](../internal/overview.md) owns implementation inventory,
    configuration model, with only the flexibility needed to express genuine
    application requirements?
 
-Both criteria must hold. ConfigKit should not preserve divergent consumer
-conventions merely for compatibility, nor impose restrictions that prevent useful
+Both criteria must hold. Minimize consumer loading and processing code subject
+to the consistency, presence, and provenance principles above, rather than
+trading those principles for consumer-specific shortcuts. ConfigKit should not
+preserve divergent consumer conventions merely for compatibility, nor impose
+restrictions that prevent useful
 domain information from being expressed. Prefer a canonical representation using
 existing primitives before adding another mechanism. Extend the model when a
 concrete requirement cannot be accommodated cleanly, not simply because a
@@ -51,7 +97,49 @@ boundaries, and keep simple consumers independent of optional composition
 features. The [consumer adoption guide](../consumers/README.md) demonstrates the
 implemented surface against representative simple and complex shapes.
 
-## Responsibility boundaries
+## Simplicity, Defaults, And Explicit Controls
+
+These principles apply throughout ConfigKit: to configuration syntax, the public
+Go API, defaults, and feature design. Evaluate their effects on both configuration
+authors and consuming applications.
+
+**Keep straightforward workflows straightforward.** Common configuration
+workflows should require little configuration text and minimal API setup.
+Support more demanding workflows through explicit, composable options, without
+requiring simple consumers to understand or configure those options.
+
+**Prefer simple, permissive defaults with explicit controls.** Choose predictable
+defaults that minimize required setup and accommodate valid configurations
+without unnecessary restrictions. Provide selectable options for workflows that
+need additional capabilities, narrower behavior, or stronger requirements.
+
+Measure simplicity by what consumers must understand and maintain, including
+configuration text and application-side plumbing. Concise syntax is useful, but
+hidden inference and interacting meanings can make a shorter declaration harder
+to understand. Prefer explicitness where it makes behavior more predictable.
+A terse configuration format that requires substantial application-side
+infrastructure does not satisfy the simplicity principle.
+
+Define what permissiveness means for each behavior. Allowing an optional
+contribution to be absent or accepting valid inputs without an unnecessary
+filename restriction can reduce required setup. Invalid declarations, ambiguous
+composition, and unreadable selected sources must still produce meaningful
+errors under the applicable contracts. Permissiveness does not authorize silent
+recovery from invalid supplied configuration or weaken configuration integrity.
+
+Options should express coherent choices with understandable purposes and fit
+ConfigKit's shared configuration model. Support powerful controls when they
+serve concrete needs; these principles do not require a switch for every
+implementation behavior or conceivable consumer preference.
+
+Defaults require judgment when simplicity and permissiveness pull in different
+directions. Broader discovery can accommodate more layouts while increasing the
+implicit search scope. Evaluate such tradeoffs explicitly; these principles do
+not mechanically select the broadest behavior for every option. Changes to
+existing defaults remain subject to compatibility review, and proposed behavior
+remains in the roadmap until implemented.
+
+## Responsibility Boundaries
 
 ConfigKit owns configuration input processing and the coordination of decoding,
 resolution, validation, and inspection. Applications own their configuration
@@ -59,19 +147,75 @@ types, field meanings, defaults, normalization, semantic validation, and schema
 version policy. Applications also own main-file search conventions, CLI wiring,
 remote acquisition, runtime initialization, and resource checks.
 
+The engine applies each input's mapping destination within the existing source
+assembly, applies immediate-parent-relative YAML destinations to imports, and
+retains each effective placement with source identity and inspection records;
+the root facade owns the public fields and conversions. The
+[configuration contract](../config.md#input-destinations) defines the observable
+placement behavior. [ADR 0011](../adr/0011-input-destinations.md) records input
+placement, and [ADR 0012](../adr/0012-relative-import-destinations.md) records
+the import extension.
+
 Library code must not depend on consuming applications or initialize their
 runtime resources. It must not perform implicit network access, discover
 application-specific paths, or change process-global state during ordinary `Load`.
+Import discovery is available only through declarations in selected sources:
+explicit `path`, `dir`, and `glob` selectors may search their declared scope, but
+ConfigKit does not search conventional application paths or discover a main
+configuration file implicitly. Imports and sibling inheritance share
+dependency-ready scheduling within their existing scopes; acquisition order and
+precedence-tier boundaries remain distinct. The [configuration contract](../config.md#explicit-imports)
+owns observable behavior; [ADR 0015](../adr/0015-shared-dependency-ready-ordering.md)
+and [ADR 0016](../adr/0016-independent-import-controls.md) record the decisions.
 The explicit secrets-directory bootstrap is a narrow, opt-in exception: applications
 invoke it during startup, and it reports writes, skips and partial failures. It
 must never run implicitly as part of configuration loading.
 Remote data can enter through application-supplied input without introducing a
 network provider into the library.
 
-## Public API and dependency direction
+## Loading Lifecycle
+
+Configuration loading is a synchronous, complete-on-return operation. A successful
+load provides the final configuration and inspection metadata after the configured
+processing and validation phases complete. A failed load reports diagnostics and
+returns no usable partial result. The normal application pattern is one load near
+startup; applications may explicitly initiate additional loads.
+
+ConfigKit releases its library-owned loading handles before returning and remains
+inactive until the application explicitly invokes it again. It performs no
+background watching, refresh, or mutation of published results. Caller-owned
+resources and application callback effects remain the caller's responsibility.
+The [configuration contract](../config.md#completion-and-resource-lifetime) owns
+the precise completion and ownership behavior. The proposed decision record is
+[ADR 0008](../adr/0008-complete-on-return-loading.md).
+
+### Accepted Processing Order
+
+Complete source composition and precedence, including selected overlays, before
+sibling inheritance. Then apply typed defaults and decode the configuration.
+Apply explicit CLI overrides to that configuration and check their types before
+running application resolution. Effective secret lookup and final validation
+follow resolution, before publication. [ADR 0018](../adr/0018-cli-after-structural-resolution.md)
+records this ordering.
+
+CLI overrides are the highest-priority input to application resolution. Supplied
+sequences replace whole sequences, and mappings combine recursively. Inheritance
+and typed defaulting do not run again after CLI application. CLI paths and values
+must not introduce structural directives; earlier source, inheritance, and
+decoding failures remain errors even if a CLI assignment would replace the value.
+Semantic validation assesses the final configuration after overrides and resolution.
+
+Applications own flag parsing and domain resolution. ConfigKit owns override
+application, validation, and source attribution. Resolution receives CLI values
+with their explicit-presence and source information. Absence-based defaults
+preserve explicit values; recorded normalization and derivation can still change
+them. Typed defaults and application resolution retain their own assignment
+semantics, independently of collection operations during source assembly.
+
+## Public API And Dependency Direction
 
 Keep one small, coherent public surface for the supported lifecycle. Simple
-single-document consumers must not need import, profile, or provider machinery.
+single-document consumers must not need import, overlay, or provider machinery.
 Do not expose parser-specific nodes, errors, or extension hooks through the
 public API. Parser integration belongs behind a private boundary, as recorded
 in [ADR 0002](../adr/0002-own-yaml-semantics.md).
@@ -94,25 +238,62 @@ Dependencies must preserve those boundaries; package splits and abstractions
 must earn their complexity. Do not impose a full application or hexagonal
 framework on this library.
 
-## Configuration integrity
+## Configuration Integrity
 
-Composition and intentional overlay are distinct operations. Composition must
-not silently become last-wins override through file ordering. The decision and
-alternatives are recorded in [ADR 0001](../adr/0001-separate-composition-and-overlay.md).
+Within a configuration tier, independent contributions use strict composition.
+Shared mappings combine recursively. Sequences with an applicable identity
+combine distinct records and reject duplicate identities, including equal
+records. Other overlapping non-mapping values remain conflicts, including equal
+values. Conflicting identity or operation declarations among peers also fail.
+Traversal order establishes stable reporting and record order; it does not grant
+precedence to peers. [ADR 0001](../adr/0001-separate-composition-and-overlay.md)
+records the distinction between strict composition and intentional overrides.
+
+The main document and its imports form a strictly composed tier. Explicit
+underlay groups compose strictly within each group, then combine from lowest to
+highest beneath the main tier. Between tiers, mappings combine recursively and
+retain omitted lower-priority children. Sequence values follow collection policy;
+other present values retain the established replacement rules. Source acquisition
+kind does not determine precedence. Input-specific constraints, including version
+ownership and path protections, remain part of the configuration contract.
+Consistent combination semantics do not require identical acquisition or lifetime
+behavior from every source kind.
+
+Built-in sibling `extends` relationships resolve after source precedence in the
+parser-independent document representation. Preserve existing mapping-sibling
+scopes and allow an identified sequence to provide a sibling scope for its
+records. A record may inherit from one parent in that sequence, selected by
+complete identity. The child supplies and keeps its own identity while inheriting
+the parent's other settings. Parents remain ordinary records subject to normal
+validation.
+
+Use the shared dependency ordering and preserve assembled sequence order.
+Resolve enclosing relationships before nested destination-local relationships.
+Mappings combine recursively; sequences use the inheriting collection's policy.
+Whole-record upsert replaces records during combination; explicit inheritance
+uses the records that remain. Inheritance remains built-in and creates no
+acquired source.
+[ADR 0009](../adr/0009-sibling-inheritance.md) records the
+original scope and continuing ownership decisions, with sequence combination
+and identified-sequence inheritance defined by
+[ADR 0017](../adr/0017-collection-sequence-policy.md) and CLI ordering by
+[ADR 0018](../adr/0018-cli-after-structural-resolution.md).
 
 ConfigKit must define supported YAML semantics explicitly rather than inherit
 parser defaults. Processing order and diagnostics must be deterministic for the
 same inputs and options. Invalid input must not produce a successful usable
 configuration. Do not run application hooks on invalid typed configuration.
 
-User-controlled precedence is application defaults, composed base, selected
-profile, then explicitly supplied CLI overrides. Applications own flag parsing;
-ConfigKit owns overlay semantics and provenance. Domain resolution follows these
-inputs and must remain recorded, not become a hidden override channel.
+User-controlled precedence is application defaults, ordered underlays, composed
+main input, selected overlay, then explicitly supplied CLI overrides.
+This priority order is distinct from processing order: typed defaults are applied
+at decoding, and application resolution follows CLI input under the lifecycle
+rules above.
 
-Preserve node-based composition until the effective input is ready for typed
-decoding. Preserve presence and source information through processing. Ordinary Go zero
-values must not be used to infer whether input was absent. Final configuration,
+Preserve node-based composition through source assembly and inheritance until
+typed decoding. Preserve presence and source information through later CLI and
+application-resolution changes as well. Ordinary Go zero values must not be used
+to infer whether input was absent. Final configuration,
 metadata, and effective inspection must agree, including application resolution
 changes; see [ADR 0003](../adr/0003-preserve-presence-and-resolution-history.md).
 Safe historical values and an explanation renderer are part of the implemented contract,
@@ -122,7 +303,80 @@ See [ADR 0005](../adr/0005-explicit-overrides-and-safe-explanation.md).
 Detailed syntax and operation tables belong to the
 [current configuration contract](../config.md).
 
-## Ownership and confidentiality
+The frozen inspection boundary also supports comparing completed same-type
+results without reopening sources or rerunning the load lifecycle. The
+[configuration contract](../config.md#effective-configuration-diffs) owns
+comparison and disclosure behavior, the [component overview](../internal/overview.md)
+owns implementation responsibility, and [ADR 0010](../adr/0010-effective-configuration-diff.md)
+records the accepted decision.
+
+## Collection Identity And Combination
+
+ConfigKit owns a common collection policy for imported records, inline sequences,
+selected overlays, and sibling inheritance. Configuration authors may declare
+identity and combination rules; applications may register authoritative identity
+constraints. Configuration declarations must agree with application constraints.
+Keep source placement explicit and separate from policy, retaining the placement
+decision in [ADR 0014](../adr/0014-sequence-placement-and-composition.md).
+[ADR 0017](../adr/0017-collection-sequence-policy.md) owns the rationale for
+collection policy and its relationship to application authority.
+
+Gather declarations for the relevant destination and source or item-local scope
+before strict composition. Applicable identity from accumulated lower tiers must
+be available when composing higher-tier peers. Higher-tier declarations cannot
+repair invalid lower-tier composition. Preserve the distinction between source
+acquisition, dependency-ready scheduling within scopes, and precedence order.
+
+Strictly compose each tier's contributions before applying one operation for the
+collection against its lower value. Support whole-sequence replacement,
+duplicate-rejecting append, and whole-record replacement by identity (`upsert`).
+Append and upsert require identity; do not infer it from record fields or position.
+Duplicate identities among peers fail under every mode. Source precedence and
+inheritance use the same operations and deterministic ordering rules. Whole-record
+upsert replaces the matching record's fields together; recursive record merging
+and unkeyed sequence concatenation are outside this operation model.
+
+An explicit mode persists at its applicable destination until a higher explicit
+mode changes it. With no explicit mode, effective identity selects upsert and
+its absence selects whole-sequence replacement. Inferred defaults do not become
+persistent instructions. The inheriting destination chooses its operation; copying
+a parent collection does not automatically copy its mode to that destination.
+
+Track YAML identity and explicit modes according to their separate lifetimes.
+Actual whole-value replacement discards lower YAML identity while preserving
+application constraints and any new identity declaration. Item-local policies
+belong to their records: retaining a record retains its rules, and discarding it
+discards those rules. Policies independently applicable at a destination remain
+in effect. Inheritance can copy YAML identity subject to agreement and replacement
+rules; application registrations continue to constrain their registered paths.
+Numeric positions alone do not establish policy ownership.
+
+Allow declarations without accompanying records. A key declaration validates a
+retained sequence immediately or remains pending in its owning scope when no
+sequence exists. A mode-only declaration affects later combinations without
+recombining earlier tiers. Neither declaration alone clears a collection, and
+declaring replace alone does not authorize changing a retained collection's
+identity. These rules do not independently impose path-existence or shape
+requirements.
+
+Enforce YAML identity during source composition, precedence, and identified
+inheritance operations. Application registrations additionally constrain final
+effective sequences after defaults, CLI overrides, and application resolution.
+The application schema and validation own existence, nullability, and allowed
+shape. The final identity guarantee covers actual sequences on successful load;
+it does not cover every intermediate callback state or later caller mutations.
+YAML-only declarations carry no such final guarantee through defaults, CLI
+assignments, or application resolution.
+
+Preserve source attribution, presence, history, sensitivity, and resource context
+as records survive, move, or are replaced. Identity diagnostics use paths and
+available origin information without exposing identity values. The guarantee
+requires no new mutation-history system or exact reconstruction of causes for
+records supplied by defaults or callbacks. Exact identity equality, declaration
+syntax, presence cases, and operation tables belong in the configuration contract
+when implemented; the roadmap owns those details until then.
+
+## Ownership And Confidentiality
 
 Returned configuration and metadata must not share mutable state across loads.
 ConfigKit must not mutate a result after returning it. Ordinary exported Go
@@ -143,12 +397,12 @@ forwarded blindly when they can include raw input. Do not log configuration
 implicitly. Applications remain responsible for classifying their secrets and
 for safe messages supplied by their hooks.
 
-## Scope boundaries
+## Scope Boundaries
 
 Do not introduce automatic directory discovery, network providers, arbitrary
 merge strategies, additional configuration languages, or application runtime
 management without concrete consumer evidence and an explicit design decision.
-The [deferred scope](../roadmap/deferred.md#explicit-non-goals-and-deferred-scope)
-owns the deferred-feature inventory. Significant boundary
-changes require an ADR and corresponding policy/contract updates; implementation
-convenience alone is not a reason to broaden the public API.
+Record significant decisions to exclude capabilities, and their rationale, in
+ADRs. Significant boundary changes require an ADR and corresponding
+policy/contract updates; implementation convenience alone is not a reason to
+broaden the public API.

docs/policy/documentation.md

@@ -1,125 +1,179 @@
 # Documentation Policy
 
-**Base Policy Version:** 1.0
+**Base Policy Version:** 1.11
 
-**Last Updated:** 2026-09-06
+**Last Updated**: 2026-10-16
+
+**Upstream Source:** [Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/docs/policy/documentation.md)
 
 ## Purpose
 
-Assign each documentation topic one canonical owner so ConfigKit's documentation
-remains accurate and useful to consumers, maintainers, and implementation agents.
+This policy assigns each documentation topic to one canonical owner. Its goal is
+to keep ConfigKit's documentation accurate, concise, discoverable, and resistant
+to drift for consumers, maintainers, developers, and coding agents.
 
-## Core rules
+## Core Rules
 
-### One canonical owner
+### One Canonical Owner
 
-Each authoritative fact belongs in one owning document or Go declaration's
-reference documentation. Other documents may summarize and link to that owner;
-they must not maintain competing definitions of commands, API contracts, YAML
-semantics, defaults, paths, diagnostics, or compatibility guarantees.
+Each authoritative fact belongs in one document. A non-owning document may give
+a short, stable summary for orientation, but it must link to the canonical owner
+for volatile details.
 
-### Policies, current behavior, and future work
+Volatile details include commands, flags, configuration fields and defaults,
+module keys, schemas, file names, paths, status codes, retry behavior, and
+runtime guarantees. If readers could reasonably treat a statement as a
+contract, maintain it only in the owning document.
 
-- Policies state currently binding contributor rules, including constraints on
-  future implementation. They do not imply that features already exist.
-- Current API/configuration contracts, consumer guidance, and component
-  documentation describe implemented behavior only, including partial boundaries.
-- ADRs record accepted decisions and rationale. Acceptance is not evidence of
-  implementation; execution status belongs in roadmaps.
-- Roadmaps own unimplemented targets, deferred work, and execution status.
+### Current And Future Behavior
 
-Future implementation plans own their task order, acceptance gates, and progress;
-`docs/roadmap/deferred.md` owns deliberate exclusions, not execution instructions.
-As behavior lands, move its normative definition into the current contract and
-replace the roadmap definition with a link. Retire completed roadmap material
-as part of release preparation; ADRs preserve significant decision history.
+Outside `docs/roadmap/`, documentation describes implemented behavior only.
+Partial features may be described only to their implemented boundary.
 
-### Audience and examples
+Architectural constraints may govern future changes. Describe current structure
+from implemented behavior.
 
-Consumer documentation explains how to use the library and what consumers own.
-Internal documents explain mechanisms and dependencies. Keep implementation
-mechanics out of consumer guidance unless they affect a meaningful choice.
+ADRs may record an accepted architectural decision before implementation. The
+roadmap owns implementation status and sequencing until the decision is
+implemented. Current architecture, user, operator, integration, and internal
+documentation are updated when the behavior lands.
 
-Complete copyable artifacts belong in `examples/`; Go API examples may live in
-executable Go example tests beside the relevant package. Prose should use the
-smallest useful excerpt and link to the maintained artifact. Examples must be
-secret-free and checked against implemented behavior. Clearly label illustrative
-future syntax in roadmaps; do not present it as a runnable current example.
+`docs/roadmap/future.md` owns potential enhancements and their open questions;
+task-specific plans own execution order and progress. Record significant
+architectural exclusions and their rationale in ADRs. Do not maintain a roadmap
+inventory of rejected proposals. Completed plans retire during release preparation.
 
-### Security and privacy
+### Audience And Detail
 
-Do not include real credentials, private keys, environment dumps, sensitive input,
-or private infrastructure details unless intentionally public. Document secret
-handling, not secret values.
+Write for the document's stated audience and include only the detail needed for
+its owned topic. Keep user and operator docs focused on using and operating the
+application. Developer docs should link to user-facing and external contracts
+for their definitions.
 
-## Canonical ownership
+### Examples
 
-| Topic | Owner | Boundary |
-| --- | --- | --- |
-| Product orientation | `README.md` | Current status, purpose, navigation, and a minimal working example |
-| Contributor entry point | `docs/development.md` | Reading guide and actual baseline validation commands |
-| Architecture rules | `docs/policy/architecture.md` | Responsibilities, dependency constraints, invariants, and scope boundaries |
-| Documentation rules | `docs/policy/documentation.md` | Ownership and document lifecycle |
-| Testing rules | `docs/policy/testing.md` | Risk-based sufficiency, test boundaries, and maintenance principles |
-| Release rules | `docs/policy/release.md` | Compatibility, integrity, gates, immutability, and exceptions |
-| Release procedure | `docs/release.md` | Exact maintainer commands, prerequisites, supported targets, and verification |
-| Exported Go API | Go documentation beside declarations | Symbol contracts, errors, ownership, and lifecycle API behavior |
-| YAML and resolution semantics | `docs/config.md` | Syntax, composition, precedence, conversion, secret references/bootstrap, and explanation semantics; Go reference docs link here for shared rules |
-| Consumer walkthroughs | `docs/consumers/`, when useful | Task-oriented usage and consumer responsibilities, not duplicate API reference |
-| Component inventory | `docs/internal/overview.md` | Implemented components and links to their owners |
-| Component mechanics | Other `docs/internal/` files, when needed | Detailed implementation behavior and change recipes |
-| Architectural decisions | `docs/adr/` | Context, decisions, alternatives, consequences, and supersession history |
-| Deferred scope | `docs/roadmap/deferred.md` | Deliberate exclusions and the conditions for reconsidering them |
-| Future implementation plans | Task-specific files under `docs/roadmap/` | Unfinished work, task order, acceptance gates, and progress; completed plans retire to Git history |
-| Historical release notes | `docs/releases/` | Version-specific historical summaries required by release policy |
-| Complete examples | `examples/` and executable Go examples | Maintained artifacts; prose links instead of duplicating full examples |
+Complete copyable files belong in `examples/`. Documentation may use the
+smallest illustrative snippet needed to explain its owned topic, but should link
+to maintained examples for complete copyable files.
 
-A listed owner is not a requirement to create an empty file before it has useful
-content. Create API, consumer, example, and component documents when the corresponding behavior
-exists and there is useful content to own. Link only to existing files; use plain
-code paths for planned files. Unvalidated candidate and publication work has the
-explicit temporary exception recorded in the release policy.
+Executable Go API examples may live beside the relevant package.
 
-ConfigKit has no owned HTTP service, application CLI, or runtime operations
-interface. Do not create those reference documents for consuming applications.
+Examples must be valid, secret-free, and tested where practical. Commands and
+configuration used in documentation should match the application.
 
-## Boundary rules
+### Security And Privacy
 
-Architecture owns governing constraints; ADRs explain why; contracts describe
-implemented semantics; internal documentation explains how. A roadmap may link
-to all of them but must not become a second current reference.
+Documentation and examples must not contain real credentials, private keys,
+private environment dumps, sensitive source material, or private infrastructure
+details unless intentionally public. Document secret-handling mechanisms, not
+secret values.
 
-Release policy owns guarantees; the release procedure owns commands. A release
-note cannot redefine a compatibility contract or replace a current-documentation
-update. The note at an immutable tag is that release's historical record.
+## Canonical Ownership
 
-This policy owns what documentation may disclose. Architecture owns security
-invariants. Current configuration and Go API contracts own secret-classification
-and rendering behavior when implemented; internal docs own their mechanisms.
+| Topic | Canonical owner | Owned content | Content owned elsewhere |
+| --- | --- | --- | --- |
+| Product orientation and minimal example | `README.md` | Purpose, current status, shortest useful Go example, and navigation, including licensing and attribution. | Full Go API and configuration contracts, release procedure, and implementation detail. |
+| Contributor entry point | `docs/development.md` | Task-oriented reading guide, contributor orientation, baseline validation commands, and canonical links. | Package inventory, architecture rules, subsystem behavior, detailed change recipes. |
+| Application architecture policy | `docs/policy/architecture.md` | System shape, normative ownership, dependency direction, boundaries, invariants, safety properties, and non-goals. | Concrete package inventory, implementation mechanics, contributor procedures, decision history, future work. |
+| Policy-set provenance and upstream navigation | `docs/policy/meta.md` | Policy-set origin, local authority, adoption baselines, and upstream adoption and update links. | Task-specific reading triggers, detailed adoption procedures, and formal attribution notices. |
+| Agent policy | `docs/policy/agents.md` | Proportionate, idiomatic implementation; rules against inventing requirements, restrictions, guarantees, or policy. | Application contracts, architecture, testing policy, and contributor procedures. |
+| Documentation policy | `docs/policy/documentation.md` | Documentation ownership, audience boundaries, maintenance rules, and ADR/document lifecycle. | Application architecture or product behavior. |
+| Testing policy | `docs/policy/testing.md` | Test philosophy, risk sufficiency, boundaries, doubles, coverage, regressions, and test maintenance. | Subsystem behavior, application contracts, subsystem-specific test inventories, and implementation plans. |
+| Release policy | `docs/policy/release.md` | Versioning, compatibility, release integrity and immutability, release documentation, proportionate gates, publication controls, completion, and required versus optional checks. | Repository-specific commands, product installation guidance, CLI version semantics, historical release summaries, and CI implementation detail. |
+| Open source licensing and attribution policy | `docs/policy/oss.md` | License defaults and author choices, COPYRIGHT and attribution formats, license/notice links, discoverability, maintenance, and concrete exceptions. | Component entries, license texts, release commands, and dependency security review. |
+| Project copyright and license grant | `COPYRIGHT` | Project identity, copyright holders and years, SPDX expression, code grant, separately licensed document scope, warranty disclaimer, and third-party statement. | Complete license texts and upstream component attribution. |
+| Project license texts | `LICENSE` | Complete project and incorporated companion license texts, with required project-specific fields. | General licensing policy and upstream component attribution. |
+| Upstream component attribution | `docs/oss/attribution.md` | House Format entries for dependencies and adapted documents: identity, material scope, copyright, license, notices, and source links. | General policy, exception reasoning, and complete standard license texts. |
+| Standard license texts | `docs/oss/licenses/` | Complete, verbatim standard license texts shared by applicable entries, including copied documentation. | Project-specific upstream notices, attribution prose, and policy interpretation. |
+| Repository-specific release procedure | `docs/release.md` | Maintainer prerequisites, candidate preparation, validation and publication commands and automated guards, asynchronous deliverables, delivery status, optional consumer verification, and recovery. | General release requirements, product installation guidance, historical release summaries, and current compatibility contracts. |
+| Exported Go API | Go documentation beside declarations | Symbol contracts, errors, ownership, and lifecycle behavior. | Shared YAML semantics and complete examples. |
+| Configuration contract | `docs/config.md` | YAML syntax, composition, precedence, schema, defaults, resolution, validation, secrets, and explanation semantics. | Go symbol contracts, complete examples, and implementation detail. |
+| Consumer guidance for the public Go API | `docs/consumers/` | Task-oriented public Go API usage, minimal examples, and consumer responsibilities. | Go symbol contracts, configuration semantics, and internal implementation detail. |
+| Inventory of implemented internal packages | `docs/internal/overview.md` | Implemented packages and components, responsibilities, and links to focused internal docs. | Normative architecture, contributor reading policy, external contracts. |
+| Behavior and architecture of internal packages | Other files under `docs/internal/` | Implementation flow, collaborators, state transitions, package-local guarantees and failures, and relevant tests. | Global architecture invariants, configuration definitions and defaults, external schemas, operator procedures. |
+| Architectural decision records | `docs/adr/` | Significant decisions, context, alternatives, rationale, consequences, and supersession history. | Current behavior reference, implementation status, task sequencing. |
+| Potential future features | `docs/roadmap/future.md` | Candidate enhancements and open questions for future consideration. | Rejected proposals, architectural decisions, task execution, and current behavior. |
+| Future implementation plans | Task-specific files under `docs/roadmap/` | Unfinished work, sequencing, acceptance, and progress; completed plans retire during release preparation. | Implemented behavior and ADR rationale. |
+| Release-note history | `docs/releases/` | One checked-in summary per source release under the procedure; the note at its immutable tag is the release record. | Current commands, behavior, contracts, and compatibility definitions. |
+| Complete examples | `examples/` and executable Go examples | Maintained copyable artifacts and Go API examples. | Configuration reference, symbol contracts, and prose explanation. |
 
-## Architectural decision records
+Create supporting documents for existing interfaces and responsibilities when
+substantive content is available.
 
-Use sequential filenames, such as `0001-separate-composition-and-overlay.md`,
-and the lightweight Nygard structure: title, status, date, context, decision,
-alternatives considered, and consequences.
+Do not create empty owners.
 
-Accepted decision content is immutable. A changed decision requires a new ADR
-and a superseded status/link in the old one. Navigation links may be repaired
-when specification ownership moves without rewriting historical rationale.
-Rejected architectural alternatives belong in ADRs; deferred or rejected product
-features belong in roadmaps. Do not create an ADR for every routine choice.
+## Boundary Rules
 
-## Maintenance and verification
+### Orientation
 
-Update the owning current documentation in the same change that implements or
-changes behavior. When ownership moves, remove the old definition and preserve
-useful navigation links.
+Link among orientation, architecture, and internal documents; maintain each
+package or behavior description in its canonical owner.
+
+### Go API And Configuration
+
+Go declaration documentation owns exported symbol behavior. The configuration
+contract owns shared YAML and resolution semantics. Consumer guidance explains
+tasks and links to those contracts.
+
+### Releases
+
+The release policy defines requirements and invariants that every release must
+satisfy. The repository-specific release procedure defines the commands and
+workflow maintainers use to satisfy that policy. Release notes record history
+at immutable tags. Maintain current behavior in its canonical documentation,
+update it with the corresponding behavior changes, and keep release procedures
+consistent with adopted policy.
+
+### Project Licensing And Upstream Attribution
+
+README and release documents link to the canonical licensing and attribution
+materials under the [OSS policy](oss.md). The release procedure owns URL
+construction and publication mechanics.
+
+### Contracts And Implementation
+
+Go API and configuration documents define externally observable behavior.
+Internal documents explain implementation and link to the owning contract.
+
+### Security Topics
+
+This policy owns what documentation and examples may contain. Architecture owns
+application security invariants. The configuration and public API contracts own
+secret-classification, bootstrap, and rendering behavior. Internal docs own
+implementation mechanisms only.
+
+## Architecture Decision Records
+
+Use sequentially numbered ADR filenames such as
+`0001-record-architecture-decisions.md`. Follow the lightweight Nygard format:
+
+1. title;
+2. status;
+3. date;
+4. context;
+5. decision;
+6. alternatives considered;
+7. consequences.
+
+Treat the decision content of an accepted ADR as immutable. When a decision
+changes, create a new ADR and update the earlier ADR's status to superseded.
+Rejected architectural alternatives and significant architectural exclusions
+belong in ADRs. Roadmaps preserve plausible future work, not rejected product
+ideas. Routine choices and rejected suggestions need no ADR unless they
+represent a significant architectural decision.
+
+## Maintenance
+
+When behavior changes, update its canonical owner in the same change. If
+ownership moves, remove the old definition and replace it with a link where
+navigation remains useful.
 
 Before completing documentation work:
 
-- verify claims against the implementation or clearly identify future design;
-- check changed local links, paths, fragments, examples, and commands;
-- preserve the distinction between policy, decisions, target behavior, and status;
-- check canonical ownership and remove stale or duplicate definitions;
-- inspect formatting and whitespace;
-- ensure no sensitive data was introduced.
+- verify affected behavior and examples;
+- check commands, flags, fields, defaults, schemas, and paths against their
+  implementation;
+- keep unimplemented behavior in the roadmap, subject to the ADR exception;
+- remove stale references and validate links;
+- confirm that non-owning documents summarize and link rather than redefine;
+- confirm that no secrets or sensitive private data were added.

docs/policy/meta.md

@@ -0,0 +1,27 @@
+# Policy Set Provenance
+
+**Base Policy Version:** 1.0
+
+**Last Updated**: 2026-09-28
+
+**Upstream Source:** [Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/docs/policy/meta.md)
+
+## Origin And Local Authority
+
+This repository's policy set is adapted from
+[Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy).
+The adopted local policies govern work in this repository. Upstream revisions
+become local policy through deliberate adoption, preserving local decisions.
+Each policy's `Upstream Source` identifies its adopted baseline.
+
+## Adoption And Updates
+
+For policy adoption, upstream revision reviews, and updates, consult the
+[upstream adoption and update guide](https://gitea.maximumdirect.net/eric/policy/src/branch/main/docs/adoption.md).
+That guide and the reference example collection are maintained upstream and
+accessed through links. Keep discovery links pointed at the current upstream
+project and guide; keep adopted-source and example links pinned to commits.
+
+The local [developer guide](../development.md) routes tasks to applicable
+policies. Formal credit and licensing notices for the copied policy material
+belong in the local [attribution document](../oss/attribution.md).

docs/policy/oss.md

@@ -0,0 +1,217 @@
+# Open Source Licensing And Attribution Policy
+
+**Base Policy Version:** 1.8
+
+**Last Updated**: 2026-09-28
+
+**Upstream Source:** [Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/docs/policy/oss.md)
+
+## ConfigKit Applicability
+
+ConfigKit's selected grant is LGPL-3.0-or-later, recorded in
+[COPYRIGHT](../../COPYRIGHT), with the complete LGPLv3 and incorporated GPLv3
+texts in [LICENSE](../../LICENSE). The [release procedure](../release.md)
+owns source-only Go module publication through Gitea. No separate hosted release
+object or license asset is required for this channel.
+
+## Purpose And Scope
+
+This policy defines project licensing conventions and a repeatable House Format
+for upstream attribution and license notices. It covers the project's own code
+and upstream components incorporated into or distributed with the application
+or library, including relevant transitive dependencies and copied code,
+documentation, or assets.
+Determine treatment from the component's terms and actual use.
+
+Consulting development guidance or using tooling alone does not require an
+attribution entry. Copying or adapting documentation for inclusion in the
+repository is different: preserve its applicable attribution and license terms.
+A material's role as development guidance does not exempt copied documents.
+Apply any express permissions for attribution-free reuse of example formatting.
+
+The baseline is a publicly available source repository with releases published
+alongside it. The format is independent of language, forge, and build system.
+It does not require attribution for every development tool, establish license
+compatibility, or resolve every obligation arising from use or distribution.
+
+## Canonical Locations
+
+The [documentation policy](documentation.md) assigns ownership:
+
+- `COPYRIGHT`: project name, copyright statements, SPDX expression, grant,
+  warranty disclaimer, and third-party statement.
+- `LICENSE` contains the complete, unmodified project license text, including
+  incorporated companion text. Dependency attribution does not relicense the project.
+- `docs/oss/attribution.md`: canonical component entries.
+- `docs/oss/licenses/`: complete, verbatim standard and applicable companion texts.
+
+Use recognizable license filenames, such as `Apache-2.0.txt` or `LGPL-3.0.txt`.
+A shared standard text may serve multiple entries. Keep explanations outside
+license texts and preserve distinct versions and exceptions.
+
+## Upstream Reference Examples
+
+Use the commit-pinned upstream examples to create actual project notices.
+Retain their links and keep the example collection upstream; keep applicable
+license texts locally as specified above. The
+[standard license-text collection](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/docs/oss/licenses/README.md)
+identifies the available reference texts.
+
+## Project Licensing
+
+Unless the author specifies otherwise, use **GPL version 3 or later**
+(`GPL-3.0-or-later`) for applications and **LGPL version 3 or later**
+(`LGPL-3.0-or-later`) for libraries. An author's explicit choice takes precedence:
+for example, a one-off script may use MIT. Such a choice requires no exception
+rationale, additional approval, or compensating controls under this policy.
+
+These defaults guide new license selection. Adopting this policy does not
+relicense existing code or override third-party terms. Preserve an existing
+license unless the author deliberately changes it. For mixed-purpose projects,
+follow the author's selected scope rather than inferring different licenses for
+individual packages.
+
+Use the upstream [application COPYRIGHT example](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/examples/licensing/application/COPYRIGHT)
+or [library COPYRIGHT example](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/examples/licensing/library/COPYRIGHT)
+to create the project’s root `COPYRIGHT` file.
+Use verified project names, contribution years, and copyright holders.
+For another selected license, adapt the grant, SPDX expression, disclaimer, and
+text references together to reflect the author's selected grant.
+
+For GPLv3 projects, copy the complete GPLv3 text into `LICENSE`. For LGPLv3
+projects, place the complete LGPLv3 text followed by the complete GPLv3 text in
+`LICENSE`; the LGPL incorporates the GPL terms. Keep the texts unmodified and
+place project-specific statements in `COPYRIGHT`. For the default grants, the
+grant and SPDX expression in `COPYRIGHT` must expressly identify `or-later`;
+follow the author’s choice if a different grant is selected.
+
+For other licenses, provide their complete text and any required project-specific
+fields; the standard reference collection remains unchanged. Keep `COPYRIGHT`,
+`LICENSE`, SPDX expressions, and README descriptions consistent with the actual
+grant. This convention introduces no blanket requirement for per-file headers;
+preserve notices or add them where the applicable terms require them.
+
+Identify separately licensed documentation or assets in the project grant, with
+links to their attribution and license information. An application's default
+code license does not override the terms of incorporated documentation.
+
+## Upstream Attribution House Format
+
+Begin `docs/oss/attribution.md` with `# Open Source Attribution Notices`,
+followed directly by component entries. These comprise the complete document.
+
+Each entry presents these five fields in order, with blank lines between them:
+
+1. **Project name** as a level-two heading. An optional identification paragraph
+   may follow, such as `Selected module: MODULE at VERSION.` with the module and
+   version in backticks. For copied documentation, a concise `Material:` sentence
+   may identify the affected files and whether they were adapted or copied
+   without modification. A factual qualifier such as `(transitive dependency)`
+   is permitted. Limit optional identification prose to factual component and
+   material scope.
+2. **Upstream copyright statement or statements**, preserving their wording.
+   Put multiple statements in separate paragraphs. Include this field when the
+   selected upstream source supplies an applicable statement; otherwise, omit it.
+3. **SPDX license identifier or expression**, when applicable, using
+   `SPDX-License-Identifier: EXPRESSION`.
+4. **License and notice links**, beginning `License and notices:`, separated
+   by ` | `, and ending with a period. Label local standard texts by license
+   name, such as `MIT` or `Apache 2.0`. Label upstream files by their actual
+   filenames, such as `COPYRIGHT`, `LICENSE`, `NOTICE`, or `PATENTS`.
+5. **Source repository link**, rendered exactly as `[Source](COMMIT_URL).`.
+
+Use the upstream [attribution example](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/examples/licensing/attribution.md)
+to create the project’s `docs/oss/attribution.md`.
+The example illustrates omitted fields and multiple copyrights and licenses.
+Replace the uppercase placeholders and illustrative URLs with actual component
+information. Local license links are relative to the destination
+`docs/oss/attribution.md`; copy the applicable texts into its `licenses/`
+directory. Do not copy example entries for components the project does not use.
+
+These are five fields, not a line limit. Omit the SPDX field when no applicable
+identifier exists; identify nonstandard terms through the linked text. Preserve
+licensing alternatives and exceptions, and record a selected option when
+relevant through the factual license identification. Do not invent an identifier
+or silently resolve an unsettled licensing choice.
+
+Keep entries factual. Do not add agent-authored explanations of license
+expressions, file-level license scope, research performed, retained source
+notices, or compliance conclusions. Avoid filler such as "none is invented
+here" and embellished link labels such as "upstream LICENSE", "additional
+patent grant", or "at the imported commit". Preserve actual required upstream
+content; this rule excludes editorial additions, not required notices. Policy
+interpretation and exception reasoning belong in the adopted OSS policy or an
+appropriate decision record, not the attribution document.
+
+## License Texts And Upstream Links
+
+Link standard license texts to their local copies. Link project-specific
+notices to their upstream copies. Every upstream repository or notice link must
+identify the exact commit corresponding to the
+imported version, including the selected source revision for copied material.
+
+Classify material according to its contents. An upstream `LICENSE`,
+`COPYING`, or similar file may contain project-specific notices or exceptions
+in addition to standard wording. Link both the local standard text and the
+commit-specific upstream file when needed. Make companion license texts directly
+reachable from the entry.
+
+For project-specific license and attribution material, this policy accepts
+direct links to the applicable upstream texts as satisfying reproduction,
+inclusion, or copy requirements. This is the adopted house interpretation.
+Agents must apply it without adding local-copy requirements or seeking repeated
+approval solely because a license uses those terms.
+
+This interpretation concerns notice reproduction. Concrete obligations such as
+marking modified files or supplying corresponding source receive their own
+applicable treatment. Preserve required notices already present in incorporated
+source files; the central attribution entry does not authorize stripping them.
+
+## Discoverability
+
+End the project's README with two friendly statements linking to its actual
+grant in `COPYRIGHT` and the attribution document. ConfigKit's
+[README](../../README.md) contains the two required statements: one linking to
+the project grant and one linking to attribution.
+
+Each new checked-in release document and its hosted download page must end
+with links whose visible text is `COPYRIGHT | LICENSE | ATTRIBUTION`. Use absolute
+URLs to those files at the release's immutable tag or corresponding immutable source
+revision so the links work in both locations. This release navigation is distinct
+from upstream component links, which must use the imported commit.
+
+Reuse the checked-in release document as the hosted release body where supported.
+The release procedure owns concrete URLs and publication mechanics for delivering
+the links through the selected channel. Do not maintain duplicate notice lists.
+These conventions do not by themselves require separate downloadable license
+assets, source bundles, or a different binary package format.
+
+## Maintenance And Exceptions
+
+Update project licensing statements together when the author changes the grant
+or its scope. Update upstream notices in the change that adds, removes, upgrades,
+copies, or changes the distribution of a relevant component. Check the selected source's license
+and notices rather than assuming they are unchanged. Update affected entries,
+commit links, and texts; retain historical versions through Git.
+
+Check affected statements, faithful standard-text copies, and link destinations
+as part of ordinary change review. For a future release tag, check intended file
+paths against the candidate source; the tag need not resolve remotely before
+publication. Do not add publication-time network checks, CI polling, recurring
+legal sign-offs, mandatory scanners, or a full software bill of materials merely
+to implement this policy.
+
+Use the House Format routinely. Raise an exception only for a concrete obligation
+or unresolved rights question it does not address, including any arising from a
+particular distribution channel. Record the issue and selected treatment in the
+adopted OSS policy or an appropriate decision record. Attribution entries
+contain the resulting factual identification and required material, without
+repeating that reasoning. Resolve the issue before the affected use or
+distribution without reopening settled choices.
+Common ownership of projects is neither required nor a blanket exception.
+
+Apply the formatting convention prospectively. Do not rewrite published tags,
+notes, binaries, or checksums to retrofit it. Address actual omissions deliberately
+under the applicable terms and the [release correction policy](release.md#immutability-and-correction).
+Ordinary attribution maintenance and release links are preparation work, not new
+release approval gates.

docs/policy/release.md

@@ -1,41 +1,39 @@
 # Release Policy
 
-**Base Policy Version:** 0.1.0
+**Base Policy Version:** 1.6
 
-**Last Updated:** 2026-09-12
+**Last Updated**: 2026-10-09
+
+**Upstream Source:** [Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy/src/commit/e833faddf120c789a9b45348edaf75720de16348/docs/policy/release.md)
 
 ## Purpose
 
-This policy defines the minimum requirements for selecting, validating,
-publishing, verifying, and correcting releases. Its goal is to make every
-release intentional, reproducible from an exact source revision, documented
-for consumers, and safe from retrospective replacement.
+This policy defines a proportionate workflow for validating and publishing an
+intended source revision under a fresh version. Releases should require little
+routine intervention, avoid repeated work, and preserve published contents.
 
-The repository-specific commands that implement these requirements belong in
-`docs/release.md`. Historical release summaries belong in `docs/releases/`.
-Neither document may weaken this policy.
+Adopt stronger verification or reproducibility requirements through an explicit
+project-specific decision justified by the intended audience and support needs.
 
-## ConfigKit applicability
+The repository-specific commands belong in `docs/release.md`. Historical release
+summaries belong in `docs/releases/`. The procedure must agree with the adopted
+policy and link to it rather than introduce additional release requirements.
 
-ConfigKit adopts a source-only Go module release model. The source tag and
-version-matched release note are the declared deliverables. Standalone binaries,
-container images, and separate binary/package-registry publication are not
-required. This does not waive source compilation, module hygiene, testing, or
-consumer-facing module retrieval and build verification.
+## ConfigKit Applicability
 
-The release procedure must define the module path, authoritative remote,
-supported Go versions and operating-system/architecture targets, prerequisite
-dependency acquisition, and the exact verification commands once established.
-The default tests themselves remain deterministic and offline. Additional
-publication channels require documented deliverables and verification before use.
+ConfigKit publishes a source-only Go module. Its declared deliverables are the
+immutable annotated source tag and version-matched checked-in release note.
+Standalone binaries, container images, hosted release objects, and separate
+registry publication are not required. Additional channels require a deliberate
+update to the policy and procedure.
 
-The Go API, documented YAML/resolution semantics, and structured diagnostic
-contracts all participate in compatibility review. A signature-preserving change
+The Go API, documented YAML and resolution semantics, and structured diagnostic
+contracts participate in compatibility review. A signature-preserving change
 can still break consumers. Human diagnostic wording is stable only where a
-contract expressly promises it. Minimum Go version and platform support changes
-also require an explicit compatibility assessment.
+contract expressly promises it. Minimum Go version and platform support
+changes also require an explicit compatibility assessment.
 
-### Platform support
+### Platform Support
 
 Linux is the release-validated platform. macOS has best-effort support: keeping
 it working remains desirable, but neither native macOS validation nor Darwin
@@ -44,49 +42,67 @@ does not by itself block a release. Voluntary checks and fixes are welcome; ther
 is no guarantee that each release has been validated on macOS.
 
 The [release procedure](../release.md#release-identity-and-support) owns target
-architectures and required Linux checks. This support change applies to upcoming
-releases; historical release notes and their validation records remain unchanged.
-The v0.3.0 release note must explain the change in Compatibility, with version
-selection following this policy's existing compatibility rules.
+architectures and required Linux checks. Historical release notes and their
+validation records remain unchanged.
 
-## Release Model
+## Release Model And Completion
 
-A release consists of:
+A source release identifies one exact committed source revision through an
+immutable version tag and a version-matched, checked-in release note. The
+primary release branch is `main` unless the adopted policy specifies otherwise.
+The procedure must name the authoritative publication remote.
 
-- one exact source commit reachable from the repository's primary branch;
-- one immutable version tag identifying that commit;
-- one version-matched, checked-in release note; and
-- every artifact, image, package, registry entry, or other deliverable the
-  repository declares for that release.
+A successful push of the intended tag, with its source commit available on the
+authoritative primary branch, completes the release. The source commit may be
+pushed before the tag or published together with it through an atomic push.
+The procedure must choose and document the supported approach.
 
-The repository's primary branch is `main` unless an adopted policy documents
-an exception. A release candidate must be committed and pushed to that branch
-before its release tag is published.
+ConfigKit uses sequential publication: first push the committed candidate to
+authoritative `main`, then invoke ReleaseKit once to validate that candidate and
+publish its annotated tag. The [release procedure](../release.md) names
+`origin`; ReleaseKit checks its exact effective push URL before validation and
+tag publication.
 
-The source tag is the authoritative release identity. Published deliverables
-must be traceable to the tagged source and must agree with its version. A
-repository may publish source only. When it publishes additional deliverables,
-its release procedure must enumerate them and define how each is built,
-identified, published, and verified.
+Release commands must finish at successful publication. They must not wait for
+or poll CI, require successful CI status, inspect downloadable artifacts, or
+check consumer installation or Go module availability as completion gates.
+A failed or ambiguous push requires resolving the authoritative remote state
+before declaring success or retrying; successful pushes do not require a
+separate remote read-back gate.
 
-Release automation may validate or publish a release only when its exact
-responsibility is documented. Automation must not silently change the source
-candidate, move a published tag, replace a published deliverable, or repair a
-release in place.
+Remote tag read-back and fresh consumer retrieval are optional and non-gating
+after successful publication. A failed push receives one bounded check of the
+authoritative ref; do not poll or blindly retry.
+
+Repositories may publish source only. Additional binaries, images, packages,
+checksums, or other declared deliverables may be published asynchronously from
+the tagged source. The procedure must identify those deliverables and the
+responsibility of their automation. Published deliverables must be traceable
+to that source and identify the same version.
+
+ConfigKit currently declares no additional deliverables. Confirmed publication
+of the intended source tag is the completion point. CI, macOS validation,
+hosted delivery, and consumer retrieval are outside the release gates.
 
 ## Versioning And Compatibility
 
 Stable releases use Semantic Versioning in the form `vMAJOR.MINOR.PATCH`, with
-no leading zeroes. Each version must be unused locally and in every configured
-remote or publication channel before publication begins.
+no leading zeroes. Prereleases require an explicit adopted policy defining
+their format and use.
 
-Unless an adopted policy explicitly supports and defines prereleases,
-prerelease versions are not permitted.
+ConfigKit does not support prereleases.
+
+The selected version must be unused locally and on the authoritative remote.
+Do not require queries to every configured remote or publication service.
+A version known to have already been published must not be reused, even if the
+normal guards do not inspect that channel. Adding a distribution channel
+requires a deliberate update to the policy and procedure where needed.
 
 For releases at or after `v1.0.0`:
 
 - a major release may make intentional incompatible changes;
-- a minor release may add backward-compatible behavior; and
+- a minor release may add backward-compatible behavior or raise the Go consumer
+  minimum as described below; and
 - a patch release may make backward-compatible fixes.
 
 Before `v1.0.0`, a minor release may change a documented public, operational,
@@ -94,146 +110,211 @@ configuration, data, integration, or compatibility contract when the release
 note explains the impact and required consumer action. A patch release must
 not intentionally break those contracts within its minor line.
 
-A repository with additional compatibility promises must define them in its
-adopted policy or canonical contract documentation. A release note may explain
-a compatibility change, but it must not redefine the current compatibility
-contract.
+Additional compatibility promises and deliberate exceptions belong in the
+adopted policy or canonical contract documentation. Release notes explain
+changes; they do not redefine current compatibility contracts.
 
-## Release Documentation
+## Consumer Minimum And Release Toolchain
 
-Every release made under this policy must include
-`docs/releases/<version>.md` in the tagged source tree. The filename must use
-the complete version tag, and the document must begin with a title containing
-the product name and that exact version.
+Distinguish the minimum toolchain supported for consumers from the toolchain
+maintainers use to build and validate a release. Record the consumer minimum in
+the language's standard metadata and explain its support meaning in canonical
+documentation. Define the selected release toolchain and its setup commands in
+`docs/release.md`. Toolchain selection and upgrade cadence are repository-specific
+choices.
 
-Every new release note must contain these sections:
+For Go repositories, normally declare the initial stable release of the selected
+series as the consumer minimum, such as `go 1.26.0`. Use a higher patch minimum
+when a dependency or an identified correctness fix requires it, and document
+the reason.
 
-```markdown
-# <Product> vMAJOR.MINOR.PATCH
+For release-policy purposes, successful build and release validation on the
+selected Go patch release are sufficient to presume support across that Go
+series, subject to the declared consumer minimum and concrete evidence of
+incompatibility. Investigate specific compatibility concerns as they arise.
 
-This release ...
+Use the selected release toolchain for candidate validation and release builds,
+including CI jobs performing those tasks. A newer validation patch version alone
+leaves the consumer minimum unchanged.
 
-## Summary
+A minor release may raise the Go consumer minimum, including from one Go series
+to the next, both before and after `v1.0.0`. This is an explicit exception to the
+backward-compatibility requirement for minor releases. Patch releases preserve
+the consumer minimum. Explain an increase's impact and required consumer action
+in the release note. Other incompatible changes remain subject to the versioning
+rules above.
 
-## Compatibility
+### ConfigKit Toolchain Selection
 
-## Upgrade
+For minor and major releases, select the newest stable Go series available in
+the development environment. For patch releases, retain the series recorded in
+`go.mod` and its consumer minimum. Use the latest locally available compatible
+patch in the selected series, with `GOTOOLCHAIN=local`, to build the
+module-selected ReleaseKit CLI and validate the candidate. The same Go
+executable builds the tool and runs its validation commands.
 
-## Changes
-```
+## Release Documentation And Preparation
 
-Use an explicit statement such as "No action required" when a section has no
-substantive content.
+Every new release must include `docs/releases/<version>.md` in the tagged
+source tree. Use the complete version tag as the filename and begin with a
+title containing the product name and that exact version.
 
-Release notes are immutable historical summaries. They should link to current
-canonical documentation for exact behavior. Every affected canonical document
-must be updated in the same candidate; a release note must not substitute for
-current documentation or defer required documentation changes.
+ConfigKit requires the first line `# ConfigKit <version>`. ReleaseKit's
+publication guard checks the release-note H1 for the product and version.
 
-## Candidate Preparation And Validation
+Include a short summary and any relevant compatibility impact or upgrade
+instructions. Choose sections suited to the release and keep the note
+proportionate to the change.
 
-Each repository must provide one documented release-check command or script
-that exercises the complete set of automated source-candidate gates. The
-repository-specific procedure must state all additional manual gates.
+Keep exact current contracts in their canonical documents and update affected
+documentation with the behavior change. Link to those documents from the note
+where useful. If annotated tags are used, derive their summary from the
+checked-in note.
 
-The required gates are:
+ConfigKit uses the complete checked-in note as the annotated tag message.
 
-- dependency or module hygiene appropriate to the language and ecosystem;
-- the complete deterministic, offline default test suite;
-- race, concurrency, or equivalent safety checks where the implementation can
-  be meaningfully checked for those risks;
-- static analysis and linting used by the repository;
-- production builds for every supported target that can be built in the
-  release environment;
-- formatting and whitespace checks;
-- validation of maintained examples, configuration, schemas, generated
-  sources, and documentation commands where applicable;
-- review of changed local documentation links;
-- removal of obsolete roadmap documents under `docs/roadmap/` that describe
-  completed work included in the release;
-- review for unintended files, generated output, credentials, private data,
-  and unrelated changes; and
-- any additional security, compatibility, migration, packaging, or platform
-  checks required by the product.
+Include the release footer defined by the [OSS policy](oss.md#discoverability)
+in each new release document. The release procedure defines the concrete URLs
+and publication mechanics. Prepare these links with the release note; they do
+not add a network check or approval gate. If a hosted download page is used in
+a future channel, preserve the footer there.
 
-Unavailable tooling or an inapplicable check must be addressed explicitly in
-the adopted policy or release procedure; it must not disappear silently from
-the candidate gates.
+Handle version selection, release-note content, and retirement of completed
+roadmaps during preparation. Preserve planned and deferred work and repair
+affected links when retiring roadmaps. Reuse ordinary change review for
+unintended files, credentials, and private data. Scope release-content review
+to the tagged source.
 
-Run the complete checker while preparing the candidate. Commit the release
-note and all associated code, tests, documentation, and configuration, then
-run the complete checker again against the exact commit intended for release.
-Do not publish a release from uncommitted or unvalidated content.
+## Candidate Validation
+
+Provide one publication command that validates the final committed candidate,
+checks publication state, and publishes the intended refs. A validation-only
+mode or command may be provided for diagnosis, but must not be a required step
+before publication repeats the same checks.
+
+ConfigKit uses the module-selected ReleaseKit CLI's `publish
+vMAJOR.MINOR.PATCH` command. Its `check` command is optional for diagnosis; it
+is not a required preliminary pass to publication.
+
+Run the complete deterministic, offline default test suite once for the final
+candidate, following the [testing policy](testing.md). For repositories that
+publish builds, verify the required build targets. Select those targets for
+the application's actual support commitments; native testing on each platform
+and access to particular hardware are not default release prerequisites.
+
+Prepare dependencies and build the tool before offline validation; the
+publication command is the required validation run, with no separate successful
+test or CI run required before it.
+
+The selected profile runs the complete deterministic, offline default suite
+once with the race detector and test-time vet disabled. The suite already
+includes examples, consumer fixtures, documentation links, and package
+boundaries. One full vet pass adds static checks beyond the test-time subset.
+CGO-disabled Linux amd64 and arm64 package compilation covers supported
+configurations distinct from native race compilation. macOS and newer-toolchain
+checks are optional.
+
+Additional gates must address a concrete risk that the existing checks do not
+adequately cover. Do not automatically require linting, race testing, module
+hygiene audits, documentation audits, credential scans, example validation, or
+expanded platform matrices solely because they are possible. Retain such
+checks when their distinct protection justifies their cost. A one-time migration
+check must not silently become a permanent release requirement.
+
+ReleaseKit runs its selected Go commands with readonly module mode and
+`GOPROXY=off`; acquire dependencies and build the tool before candidate
+validation. Formatting and whitespace remain ordinary development checks in
+`scripts/check.sh`.
+Review module tidiness when dependencies change, not as a release gate.
+
+Release tests must not require real credentials, mutable application services,
+paid APIs, provider probes, or inference calls. The procedure may describe dependency
+preparation separately from offline validation.
+
+Avoid duplicate validation across release phases. Successful checks for the
+same final candidate need not be repeated solely to satisfy another phase;
+any reuse must clearly apply to that exact candidate. Persistent validation
+caches are not required. If the candidate changes, validate the new candidate.
+Cheap publication-state checks may be repeated immediately before pushing
+because local or remote state may have changed during validation.
+
+Automate mechanical guards. Human review gates are strongly disfavored: add one
+only through an explicit project-specific policy decision explaining the
+concrete risk and why ordinary change review or automation is insufficient.
+Do not add routine confirmations for checks the command can perform itself.
+
+Compatibility review, release-note accuracy, completed-roadmap retirement, and
+review for unintended files or credentials remain ordinary preparation. The
+executing agent may perform them without routine human approval or separate
+manual evidence records. Publication itself requires authorization.
 
 ## Publication Guards
 
-Immediately before publication, the maintainer or release automation must
-fetch the current primary branch and tags from the authoritative remote and
-verify all of the following:
+Automate the checks needed to publish the intended source under a fresh version:
 
-- the version has the required format and is not already published;
-- the current branch is the primary release branch;
-- the worktree and index are clean;
-- the current commit is the recorded, validated candidate;
-- the authoritative remote branch points to that same commit;
-- the matching release note exists in that commit and has the required title
-  and sections;
-- all required approvals and automated gates succeeded for that commit; and
-- no local or remote tag already uses the selected version.
+- the version has the adopted format and is unused locally and on the
+  authoritative remote;
+- the worktree and index are clean on the primary release branch;
+- the source being published is the exact candidate that passed validation,
+  including its matching release note H1;
+- the remote primary branch can be advanced to that candidate without
+  discarding remote history, and is published before the tag; and
+- the tag identifies the validated candidate and only the intended refs are
+  published.
 
-Create the version tag against the guarded commit and verify its target before
-publishing it. Publish only the exact intended ref and deliverables. Broad
-operations such as `git push --tags` are discouraged because they can publish
-unreviewed refs unrelated to the release.
+ReleaseKit checks that `origin`'s effective push URL exactly matches the
+authoritative URL before candidate validation. This does not assert the
+configured fetch URL or require the release-note H1 to be the first line; the
+required `# ConfigKit <version>` first-line format remains a content rule. The
+command checks the candidate and unused version again after validation, and
+checks local state before creating the tag. Ignored local files do not block
+publication.
 
-The repository-specific procedure must make these guards executable or give
-copyable commands for performing them.
+Fetch the authoritative branch and tag state as needed for these checks.
+Do not force branch or tag updates to bypass a failed guard. Publish explicit
+refs rather than all local tags. Do not require CI success, publication-service
+availability checks, or a separate manual confirmation for each guard.
 
 ## Immutability And Correction
 
-Published version tags and published deliverables are immutable. Do not move,
-overwrite, recreate, or delete a published tag. Do not replace an artifact,
-image, package, checksum, signature, provenance statement, or release note
-under an already published version.
+Published version tags and published contents are immutable. Do not move,
+overwrite, recreate, or delete a published tag, or replace a published artifact,
+checksum, signature, provenance statement, or release note under that version.
+The checked-in note at the immutable tag is that release's historical record.
 
-If validation fails before publication, correct the candidate, rerun the
-complete gates, and repeat the publication guards. An unpublished local tag
-may be removed after confirming that it was never published or used by a
-release channel.
+If validation fails before publication, correct and commit the candidate and
+rerun the publication command. If a push fails or its outcome is ambiguous,
+inspect the authoritative remote before retrying. Remove an unpublished local
+tag only after confirming it was never published or used by a release channel.
+An unavailable remote leaves publication status unknown; do not overwrite a tag
+to resolve uncertainty.
 
-If publication or post-publication verification reveals a defect, leave the
-published release intact. Correct the defect on the primary branch, choose a
-new patch version or other semantically appropriate new version, write a new
-matching release note, and perform the complete release procedure again.
+Transient network, authentication, CI, or consumer-availability failures do not
+by themselves require a new version.
 
-A distribution channel's supported yank, deprecation, or advisory mechanism
-may warn consumers away from a defective release, but it must not replace that
-release's contents or reuse its version.
+A defect requiring changes to published source or contents requires a corrected
+commit, a new semantically appropriate version, a matching release note, and
+the release procedure again. Leave the original release intact. A channel's
+yank, deprecation, or advisory mechanism may warn consumers away from a defective
+release without replacing its contents or reusing its version.
 
 ## Repository Release Procedure
 
-Every adopting repository must maintain `docs/release.md` as the canonical
-maintainer runbook for implementing this policy. It must define, in execution
-order:
+Maintain `docs/release.md` as a short, executable maintainer runbook. It must
+identify prerequisites, the selected release toolchain, the primary branch and
+authoritative remote, version and note preparation, one publication command with
+local checks and guards, source-only deliverables, the successful-tag-push
+completion point, and partial or unknown-state recovery.
 
-1. maintainer prerequisites and required access;
-2. version selection and compatibility review;
-3. release-note creation;
-4. candidate preparation and the complete validation command;
-5. candidate commit, push, and exact commit recording;
-6. publication guards;
-7. exact tag and deliverable publication commands;
-8. authoritative remote and consumer-facing verification; and
-9. failure handling and correction.
-
-The procedure should use copyable, fail-fast commands where practical. It must
-name the primary branch, authoritative remote, tag form, published
-deliverables, supported build targets, release automation responsibilities,
-and consumer verification path. It must link to this policy instead of
-restating or redefining its requirements.
+Use copyable commands where practical. Keep optional diagnostics and consumer
+verification clearly optional. The procedure must not turn preparation work,
+ordinary change review, or post-publication investigation into additional
+release gates.
 
 ## Policy Exceptions
 
+The following historical exceptions do not change the current release workflow.
+
 For v0.1.1 only, the maintainer approved renaming the reserved `_configkit` key to
 `configkit` as an error-correcting point release because there are no current
 consumers. This intentionally changes the v0.1.0 configuration contract without a
@@ -246,6 +327,3 @@ For v0.2.0 only, the maintainer accepted the successful native macOS check at
 only. The final candidate still requires the complete Linux release checker.
 This exception waives only repeating the macOS check after those documentation
 changes; it does not waive other publication or verification gates.
-
-Candidate selection, publication, and post-publication verification remain
-mandatory steps in the [release procedure](../release.md).

docs/policy/testing.md

@@ -1,14 +1,16 @@
 # Testing Policy
 
-**Base Policy Version:** 1.0
+**Base Policy Version:** 1.3
 
-**Last Updated:** 2026-09-06
+**Last Updated**: 2026-09-28
+
+**Upstream Source:** [Standard Repository Policies](https://gitea.maximumdirect.net/eric/policy/src/commit/d84a38ebab26f048cafbe59a2246700c35d29843/docs/policy/testing.md)
 
 ## Purpose
 
 Our tests exist to make **incorrect changes expensive and correct changes cheap**.
 
-We do not optimize for test count, line coverage, exhaustive isolation, or the fewest possible tests. We optimize for sufficient confidence in important behavior while imposing as little unnecessary friction as possible on future development.
+We optimize for sufficient confidence in important behavior while minimizing friction on future development.
 
 ## ConfigKit-specific priorities
 
@@ -44,32 +46,19 @@ policy does not imply that those tools or tests already exist.
 
 ## Every test has a cost
 
-Testing is not an unqualified good. Every test imposes both an immediate cost and a continuing lifetime cost.
-
-A test must be:
-
-- written and reviewed;
-- understood by future maintainers and coding agents;
-- executed in local and CI workflows;
-- diagnosed when it fails;
-- updated when legitimate behavior changes;
-- maintained as fixtures, APIs, and dependencies evolve; and
-- removed or rewritten when it becomes redundant, brittle, misleading, or obsolete.
-
-Tests also create cognitive and architectural friction. They can constrain refactoring, duplicate policy, slow feedback loops, add noise to failures, and cause harmless implementation changes to require unrelated edits across the suite.
+Every test has an immediate and continuing cost: writing, review, execution,
+diagnosis, fixture maintenance, and updates when legitimate behavior changes.
+Tests can also constrain refactoring, duplicate policy, and add failure noise.
 
 A test is warranted only when the confidence it provides justifies these costs.
-
-Apply this cost-benefit analysis at two levels:
+Evaluate that tradeoff at two levels:
 
 1. **Per test:** What realistic defect does this test detect, how consequential would that defect be, and is that protection worth the test's lifetime cost?
 2. **Across the suite:** Does this collection provide materially more confidence than a smaller, simpler suite would?
 
-The preferred test suite is a **lean suite that provides sufficient confidence in the risks that matter, without redundant or low-value tests**. We seek sufficient confidence with the least unnecessary testing friction, not the fewest possible tests.
+Some friction is intentional. Tests should make dangerous changes—such as breaking compatibility, corrupting data, violating security boundaries, or reintroducing subtle bugs—require deliberate review. Keep ordinary internal changes inexpensive.
 
-Some friction is intentional. Tests should make dangerous changes—such as breaking compatibility, corrupting data, violating security boundaries, or reintroducing subtle bugs—require deliberate review. They should not make ordinary internal changes needlessly expensive.
-
-The cost of a test is not a reason to omit testing by default. Do not cite maintenance cost abstractly. When omitting a plausible test, be able to state why the protected failure is low-risk, already covered, obvious, reversible, or cheaper to detect elsewhere. For consequential, subtle, or difficult-to-observe behavior, the presumption should favor testing.
+Evaluate maintenance cost against the specific failure being protected. When omitting a plausible test, be able to state why the protected failure is low-risk, already covered, obvious, reversible, or cheaper to detect elsewhere. For consequential, subtle, or difficult-to-observe behavior, the presumption should favor testing.
 
 ## Default testing style
 
@@ -81,6 +70,8 @@ Use a **classical/Detroit-style** approach:
 - Prefer package-level behavioral tests over tests coupled to private helpers or internal call sequences.
 - Treat exact collaborator interactions as testable behavior only when the interaction itself is a requirement.
 
+Examples of appropriate seams include clocks, randomness, subprocesses, remote APIs, object storage, email, and paid LLM calls.
+
 For ConfigKit, controlled input readers and filesystem failures are useful seams.
 Prefer actual parsing, composition, and decoding over mocks of their call order.
 
@@ -92,7 +83,7 @@ Control clocks, randomness, environment variables, and other process-global or m
 
 ## What deserves tests
 
-Prioritize tests for:
+To the extent the following features are present in the application, prioritize tests for:
 
 1. Public and package-level contracts.
 2. Domain rules and important invariants.
@@ -102,15 +93,27 @@ Prioritize tests for:
 6. Previously observed or plausible regressions.
 7. Representative integration and end-to-end workflows.
 
-A package-level contract is behavior relied upon by another package or major collaborator, not every observable detail of a package implementation.
+A package-level contract comprises behavior relied upon by another package or major collaborator.
 
-For behavior involving **data integrity, destructive operations, compatibility, security, concurrency, idempotency, or recovery**, presume that durable tests are required unless the behavior is already credibly protected at another layer.
+Choose tests for requested behavior and established contracts. Tests must not
+introduce new restrictions, guarantees, or threat scenarios merely because they
+could be asserted. Follow the [agent policy](agents.md) when deciding whether a
+behavior belongs to the contract; current implementation behavior alone is not
+sufficient authority.
+
+For established behavior involving **data integrity, destructive operations,
+compatibility, security, concurrency, idempotency, or recovery**, favor durable
+tests for realistic failures whose consequences warrant protection. Identify
+the failure and the existing contract it would violate. These categories call
+for attention, not an automatic test inventory or permission to strengthen the
+application's guarantees. Account for protection already provided at other layers.
 
 Do not add tests merely because a function, branch, or line exists. Do not add a test when the same meaningful risk is already adequately protected elsewhere.
 
 ## Choose the right test boundary
 
-Test through the narrowest stable boundary that expresses the behavior clearly.
+Choose the narrowest stable boundary that expresses the behavior clearly and
+provides durable confidence with minimal incidental coupling.
 
 This is often the package API, but it may instead be:
 
@@ -118,28 +121,15 @@ This is often the package API, but it may instead be:
 - a package-level operation when several internal collaborators jointly produce the behavior; or
 - a larger integration boundary when correctness emerges from interaction with a real dependency.
 
-Do not force all behavior through oversized end-to-end tests. Do not test every private helper merely because it exists. Choose the boundary that gives durable confidence with the least incidental coupling.
-
 ## Test behavior, not implementation
 
-A test should protect a decision, contract, or invariant—not memorialize the current implementation.
+A test should protect a selected decision, contract, or invariant.
 
-Before adding or retaining a test, ask:
+A test is suspect when it primarily detects changes to private constants or
+helpers, equivalent operation ordering, incidental formatting, equivalent
+algorithms, or internal structure that preserves behavior.
 
-> What realistic defect would this test catch?
-
-A test is suspect when its main purpose is to detect that someone:
-
-- changed an internal constant;
-- renamed or split a private helper;
-- reordered equivalent internal operations;
-- changed incidental formatting;
-- replaced one correct algorithm with another; or
-- refactored internal object structure without changing behavior.
-
-Refactoring should normally require no test edits unless the refactored structure is itself part of the contract.
-
-A test can be factually correct and still have negative value. Accurately describing current behavior is not enough; the protected behavior must be important enough to justify the future friction.
+Behavioral tests should normally remain stable through internal refactoring that preserves the contract.
 
 ## Expected effects of different changes
 
@@ -156,12 +146,13 @@ A test failing is not the same as a test needing to be edited. Many tests may co
 
 ## Separate mechanism from policy
 
-Configurable thresholds and defaults must not be duplicated throughout the test suite.
+Derive expectations from test-controlled configuration and relationships.
 
-For example, do not encode an internal concurrency limit indirectly:
+For this example, assume the selected contract rejects a request to start more
+processes than a configured limit. Do not encode its production default indirectly:
 
 ```go
-// Production policy:
+// Production default:
 const maxConcurrency = 4
 
 // Brittle test:
@@ -179,25 +170,30 @@ require.NoError(t, runner.Start(limit))
 require.ErrorIs(t, runner.Start(limit+1), ErrTooMuchConcurrency)
 ```
 
-The test should prove:
+For this selected contract, the test should prove:
 
 - the configured limit is accepted; and
 - one beyond the configured limit is rejected.
 
+Rejection is not a general requirement for concurrency control. If the selected
+contract queues excess work or limits active workers while accepting larger
+batches, test that behavior instead. Do not add rejection or a configurable
+limit solely to reproduce this example.
+
 The production default should be tested exactly only when its literal value is itself a public, operational, safety, protocol, or compatibility requirement.
 
-Apply the same rule to limits, timeouts, capacities, retry counts, and ranges: test relationships and behavior, not duplicated literals.
+Apply the same rule to limits, timeouts, capacities, retry counts, and ranges: test relationships and behavior.
 
-For concurrency limits, test both kinds of behavior when relevant:
+For established concurrency limits, distinguish these behaviors where the contract requires them:
 
-1. **Configuration enforcement:** invalid or excessive requested values are handled correctly.
+1. **Configuration handling:** values are accepted or rejected according to the selected contract.
 2. **Runtime enforcement:** observed peak concurrency never exceeds the configured limit.
 
-Use a test-controlled limit and measure the behavior relative to that limit. Do not merely assert today's default value.
+Use a test-controlled limit and measure the behavior relative to that limit.
 
-## Avoid semantic duplication across layers
+## Assign each behavior a test owner
 
-Each behavior should have a clear test owner.
+For ConfigKit, assign ownership as follows:
 
 - YAML boundary tests own syntax and policy enforcement.
 - Composition and overlay tests own their operation semantics.
@@ -209,9 +205,7 @@ Each behavior should have a clear test owner.
 
 These are responsibility boundaries, not a required package layout.
 
-Higher-level tests should not repeat every lower-level case. A single intentional policy change should not require unrelated edits across many test files.
-
-Tests that are individually reasonable may still be collectively redundant. Evaluate the marginal value of each additional test in light of the protection already provided by the rest of the suite.
+Give each layer its distinct responsibility and keep expectation updates localized to the contract being changed.
 
 ## Use test doubles deliberately
 
@@ -232,104 +226,77 @@ Mocks are appropriate when the contract includes facts such as:
 - an expensive API is called no more than once; or
 - a security audit event is emitted.
 
-Do not use mocks merely to isolate every object or reproduce the implementation's call graph.
-
 ## Go-specific guidance
 
-Use:
+Select techniques by behavior and risk:
 
 - table-driven tests for meaningful behavioral categories and boundaries;
 - `t.TempDir()` for real filesystem behavior;
-- fuzz tests for parsers, normalization, path handling, and broad input spaces;
+- `httptest.Server` for realistic HTTP interactions;
+- fuzz tests when exploring broad input spaces, such as parsers or path handling, adds useful protection beyond representative cases;
 - golden files only when the complete output is intentionally stable;
 - integration tests where correctness depends on component interaction; and
-- a small number of representative end-to-end tests.
+- a small number of representative end-to-end tests when assembled workflows expose meaningful risks not adequately covered at narrower boundaries.
 
-Avoid exact error-string assertions unless the wording is itself contractual. Prefer `errors.Is`, `errors.As`, typed errors, or structured error fields.
+Prefer `errors.Is`, `errors.As`, typed errors, or structured error fields. Use exact error-wording assertions only when the wording is contractual.
 
 At the public library boundary, prefer diagnostic codes, paths, and source
 locations over complete message snapshots. Assert wording only when contractual.
 
 Golden-file updates must require an explicit local flag. CI must not update golden files automatically, and reviewers must inspect the semantic diff before accepting an update.
 
-Keep tests readable and direct. Test helpers and fixture frameworks must earn their own maintenance cost; do not build elaborate test infrastructure for small or isolated needs.
+Keep tests readable and direct. Size test helpers and fixture frameworks to the maintenance benefit they provide.
 
 ## Coverage
 
-Coverage is a diagnostic, not a target.
+Use coverage to locate untested critical branches and unexpectedly weak packages. Evaluate test quality through meaningful behavioral protection and choose coverage according to risk.
 
-Use it to find untested critical branches and unexpectedly weak packages. Do not write low-value tests solely to increase a percentage, and do not infer test quality from coverage alone.
-
-Dense configuration semantics will often warrant higher coverage than simple input wiring. Uneven coverage is acceptable when it reflects risk.
-
-Increasing coverage is valuable only when the newly covered behavior protects a meaningful risk at an acceptable cost.
+Pure domain logic will often warrant higher coverage than CLI wiring or external adapters. Uneven coverage is acceptable when it reflects risk.
 
 ## Regression tests
 
 A bug fix should normally include a regression test that fails before the fix and passes afterward.
 
-Retain the test when the defect could realistically recur and its consequences justify the ongoing cost. Prefer the narrowest durable test of the violated contract or invariant; do not preserve accidental implementation details from the original bug.
+Retain the test when the defect could realistically recur and its consequences justify the ongoing cost. Prefer the narrowest durable test of the violated contract or invariant.
 
-Not every historical bug requires a permanent test. If the underlying design has made recurrence impossible, the test has become redundant, or a stronger invariant test now subsumes it, remove or consolidate it.
+Remove or consolidate a test when the underlying design has made recurrence impossible, its protection is redundant, or a stronger invariant test subsumes it.
 
 ## Deleting or rewriting tests
 
-Tests are maintained code, not permanent historical artifacts.
-
-Delete or rewrite a test when its maintenance cost exceeds the confidence it provides.
+Delete or rewrite tests when maintenance cost exceeds their continuing
+protective value.
 
 Strong candidates include tests that:
 
 - require updates after harmless internal changes;
-- directly assert private constants without protecting a real contract;
-- duplicate the same policy across several layers;
+- assert incidental private constants or snapshot large amounts of incidental output;
 - verify mock choreography rather than outcomes;
-- snapshot large amounts of incidental output;
-- test trivial private helpers already exercised through stable package behavior;
-- protect risks already covered more effectively elsewhere;
+- duplicate policy across layers, retest trivial helpers covered by package
+  behavior, or protect risks covered more effectively elsewhere;
 - are flaky, misleading, obsolete, or disproportionately expensive to diagnose; or
 - no longer correspond to a plausible failure mode.
 
-Several brittle tests may encode one genuine requirement. Replace them with one durable behavior-level or invariant test rather than preserving all of them.
-
-Deleting a low-value test can improve the quality of the suite by reducing noise, maintenance burden, and friction around legitimate change.
+Several brittle tests may encode one genuine requirement. Consolidate them into one durable behavior-level or invariant test.
 
 ## Reviewing a proposed test
 
-Use the following questions when the value, boundary, or durability of a proposed test is not self-evident. Significant test additions should be reviewable against them, but written answers are not required for every routine test.
+Use the following questions when a proposed test needs closer evaluation. Significant test additions should be reviewable against them; written answers are optional for routine tests.
 
-1. What realistic defect would it catch?
-2. How likely is that defect?
-3. How consequential would it be?
-4. Is the behavior already protected elsewhere?
-5. At which layer should this behavior be owned?
-6. Does the test assert a durable contract or an incidental implementation detail?
-7. Could the implementation be refactored without changing the behavior and without editing this test?
-8. What should cause this test to fail?
-9. What legitimate changes should not cause this test to fail?
-10. What ongoing maintenance, execution, and diagnostic cost will the test impose?
-11. Is there a smaller or more direct test that protects the same risk?
-
-Do not add the test when its expected lifetime cost exceeds its expected protective value.
-
-When deciding not to test plausible behavior, record or be able to explain why the risk is low, already protected, obvious, reversible, or cheaper to detect elsewhere.
+1. Which requested behavior or established contract does it protect, and what realistic defect would violate it?
+2. How likely and consequential is that defect, and is it already protected elsewhere?
+3. What is the narrowest stable boundary that can test it clearly?
+4. Would a correct internal refactor or a change to an incidental default require test edits?
+5. Does its confidence justify its maintenance, execution, and diagnostic cost, or would a smaller test suffice?
 
 ## Definition of sufficient
 
 A test suite is sufficient when:
 
-- important contracts and invariants are protected;
-- meaningful boundaries and failure modes are exercised;
-- realistic and consequential regressions are credibly protected against silent recurrence;
-- behavior involving data integrity, destructive operations, compatibility, security, concurrency, idempotency, and recovery is credibly protected;
-- important external boundaries have realistic integration coverage;
-- representative complete workflows are tested;
-- failures provide useful signal rather than redundant noise;
-- legitimate internal changes usually do not require test edits; and
+- important requested behavior and established contracts are protected against realistic, consequential failures;
+- boundaries and component interactions have coverage where they introduce meaningful risks not protected elsewhere;
+- failures provide useful, distinct signals;
+- tests usually remain stable through legitimate internal changes; and
 - additional tests would mostly repeat existing protection or preserve inconsequential implementation details.
 
-Sufficiency is a risk judgment, not a coverage percentage or test count. Reassess it as the library, its consumers, and the consequences of failure evolve.
-
-The governing rule is:
-
-> Test heavily where failure is consequential, subtle, or difficult to detect after the fact. Test lightly where failure is obvious, reversible, and inexpensive—and retain no test whose lifetime cost exceeds the confidence it provides.
+Reassess sufficiency as the application, its users, and the consequences of
+failure evolve.

docs/release.md

@@ -1,244 +1,237 @@
 # ConfigKit Release Procedure
 
 This is the maintainer runbook for ConfigKit's [release policy](policy/release.md).
-ConfigKit publishes one source-only Go module: the immutable source tag and the
-matching checked-in release note are the deliverables. There is no application
-binary, container image, generated package, or separate registry publication.
+ConfigKit publishes a source-only Go module. The immutable annotated tag and
+matching checked-in release note are the deliverables; there is no binary,
+container image, generated package, or separate registry publication.
 
 ## Release identity and support
 
 - Module: `gitea.maximumdirect.net/eric/ConfigKit`
 - Primary branch: `main`
-- Authoritative remote: `origin`, expected to fetch from
-  `https://gitea.maximumdirect.net/eric/ConfigKit.git`
-- Tags: stable Semantic Versions such as `v0.1.0`; prerelease tags are not supported
-- Minimum Go language/toolchain line: Go 1.25, tested with a supported patched
-  Go 1.25 toolchain and one newer supported toolchain
-- Release-gated build targets: Linux on amd64 and arm64
-- Best-effort targets: macOS on amd64 and arm64, under the
+- Authoritative remote: `origin`
+- Authoritative push URL: `https://gitea.maximumdirect.net/eric/ConfigKit.git`
+- Tags: stable Semantic Versions such as `v0.5.0`; no prereleases
+- Minimum Go version: the first stable patch of the selected Go series, or a
+  higher patch when required by a dependency or specific correctness fix, as
+  recorded in `go.mod`; see the
+  [consumer minimum policy](policy/release.md#consumer-minimum-and-release-toolchain)
+- Release-gated targets: Linux amd64 and arm64
+- Best-effort targets: macOS amd64 and arm64, under the
   [platform support policy](policy/release.md#platform-support)
 
-The automated checker cross-builds the Linux targets. Before release, the maintainer
-must also obtain native test, race, symlink, and permission-fixture evidence on Linux;
-cross-compilation is not native execution evidence. Native macOS checks and Darwin
-cross-builds are optional and do not gate publication.
+Release validation runs one offline race-enabled Go suite, one full vet pass,
+and CGO-disabled package compilation for both Linux targets. macOS and CI are
+not release gates; do not wait for or poll CI.
 
-## 1. Prepare access and tools
+## 1. Select Go and prepare the tool
 
-The release maintainer owns this procedure. Work in a dedicated clone with Git, Bash 3.2 or newer,
-and two supported Go toolchains available. The checker supports the stock macOS
-Bash; its filename enumeration is null-delimited, including names with whitespace. Configure private-host Git and Go module
-access through the normal credential manager before starting. Keep tokens, passwords,
-SSH material, and environment dumps outside the repository and command transcript.
-Verify access without displaying credentials:
+Use a clean checkout with Git, Bash 3.2 or newer, and a native compiler for race
+tests. Choose the release type before selecting Go. For a minor or major release,
+select the newest stable Go series available in the development environment.
+For a patch release, retain the series recorded in `go.mod` and leave its consumer
+minimum unchanged. Follow the
+[toolchain selection policy](policy/release.md#configkit-toolchain-selection).
+
+Put the selected series' latest locally available compatible patch first on
+`PATH`, ensuring it meets `go.mod`'s minimum. The same `go` executable must build
+ReleaseKit and run its child validation commands. Set local toolchain selection
+and verify it:
 
 ```sh
-git remote get-url origin
-git ls-remote origin HEAD
+export GOTOOLCHAIN=local
+go version
+go env GOVERSION
+command -v go
 ```
 
-The remote URL must equal the authoritative URL above. Select the latest supported
-patch in the Go 1.25 line and a newer supported toolchain. The values below are the
-toolchains used while this runbook was introduced; update them to supported patch
-releases during candidate preparation and record the actual versions in the release
-note or validation record.
+Only when a minor or major release advances the consumer minimum to a newer Go
+series, derive that series' initial stable release from the selected executable:
 
 ```sh
-export CONFIGKIT_BASELINE_TOOLCHAIN=go1.25.5
-export CONFIGKIT_NEWER_TOOLCHAIN=go1.26.7
-GOTOOLCHAIN="$CONFIGKIT_BASELINE_TOOLCHAIN" go version
-GOTOOLCHAIN="$CONFIGKIT_NEWER_TOOLCHAIN" go version
-GOTOOLCHAIN="$CONFIGKIT_BASELINE_TOOLCHAIN" go mod download
-GOTOOLCHAIN="$CONFIGKIT_NEWER_TOOLCHAIN" go mod download
+selected_go=$(GOTOOLCHAIN=local go env GOVERSION)
+selected_series=${selected_go#go}
+selected_series=${selected_series%.*}
+release_minimum="$selected_series.0"
+go mod edit -go="$release_minimum"
 ```
 
-Those commands explicitly acquire toolchains and dependencies. The checker itself
-sets `GOPROXY=off` and fails if its required cache is incomplete.
+If a dependency or identified correctness fix requires a higher patch minimum,
+use that required version instead and document the reason. When the series is
+unchanged, leave the minimum unchanged unless a minor or major release
+deliberately changes it under the
+[consumer minimum policy](policy/release.md#consumer-minimum-and-release-toolchain).
+For example, validating with Go 1.26.8 leaves `go 1.26.0` unchanged. There is no
+separate requirement to validate with Go 1.26.0.
 
-## 2. Select the version and review compatibility
-
-Choose an unused stable version and compare the candidate with the previous tag.
-For the first release, review the whole public surface and current contracts.
+Prepare dependencies whether or not the minimum changes:
 
 ```sh
-release_version=v0.1.0
-printf '%s\n' "$release_version" | grep -Eq '^v(0|[1-9][0-9]*)\.(0|[1-9][0-9]*)\.(0|[1-9][0-9]*)$'
-git show-ref --verify --quiet "refs/tags/$release_version" && exit 1 || :
-if git ls-remote --exit-code --tags origin "refs/tags/$release_version" >/dev/null; then
-  printf 'remote tag already exists: %s\n' "$release_version" >&2
-  exit 1
-else
-  test $? -eq 2
-fi
-git log --oneline --decorate --no-merges
+go mod download
 ```
 
-Review exported Go signatures and ownership, YAML and scalar behavior, precedence,
-diagnostic codes, minimum Go version, and platform support. Before `v1.0.0`, any
-intentional incompatible minor-release change must be explained with consumer action;
-patch releases must remain compatible within their minor line. Review the maintained
-consumer examples and the migration guide even when signatures did not change.
-
-Reassess `go.yaml.in/yaml/v4` rather than using `@latest` implicitly:
+Build the tool selected by `go.mod` into temporary storage outside the checkout.
+Ensure `$TMPDIR` names a system temporary location, not the repository. These
+commands never install another ReleaseKit version:
 
 ```sh
-go list -m -versions go.yaml.in/yaml/v4
-go list -m -json go.yaml.in/yaml/v4@latest
+releasekit_tmp=$(mktemp -d "${TMPDIR:-/tmp}/configkit-releasekit.XXXXXX")
+releasekit_bin="$releasekit_tmp/releasekit"
+go build -o "$releasekit_bin" gitea.maximumdirect.net/eric/ReleaseKit/cmd/releasekit
+go list -m gitea.maximumdirect.net/eric/ReleaseKit
+"$releasekit_bin" version
 ```
 
-If changing the explicit pin, make that a reviewed dependency change and rerun the
-complete checker, especially parser policy, limits, diagnostics, and rendering tests.
-If retaining a prerelease pin, record that decision in Compatibility.
+The module-selected executable is ReleaseKit v0.4.1. Use the built executable
+for publication to preserve its exit status and publication-state diagnostics.
+`go run` maps a nonzero child exit status to 1. ReleaseKit's child Go commands
+use the `go` found on `PATH` with `GOTOOLCHAIN=local`; keep the selected
+executable first on `PATH` through the publication command.
 
-## 3. Write the release note
-
-Create `docs/releases/$release_version.md` before selecting the candidate. It must
-begin with `# ConfigKit $release_version` and contain `Summary`, `Compatibility`,
-`Upgrade`, and `Changes` sections. Use “No action required” where appropriate. Link
-to current canonical contracts instead of copying their definitions. Do not create a
-release note until there is a real candidate and reviewed content.
-
-## 4. Validate the source candidate
-
-Run the complete checker on Linux from the repository root after dependencies have been
-acquired. It validates module manifests in a disposable copy, both supported Go
-toolchains, offline default tests, the native race detector, vet, native builds,
-CGO-disabled Linux cross-builds, formatting and whitespace (including untracked files),
-examples, package isolation, and local documentation links.
+Verify access to the authoritative remote without displaying credentials:
 
 ```sh
-./scripts/release-check.sh
+git ls-remote https://gitea.maximumdirect.net/eric/ConfigKit.git HEAD refs/heads/main
 ```
 
-`go vet` is the repository's static-analysis baseline. No additional linter is
-claimed. Links to current ConfigKit files and headings are strict. Recorded links to
-sibling consumer source are checked when those checkouts exist and are otherwise
-inapplicable in an isolated clone; their immutable revisions remain recorded in the
-extraction inventory.
+Publication requires explicit authorization. Local implementation commits do
+not authorize a push or release.
 
-Complete these manual gates and record their evidence with the candidate:
+## 2. Prepare and commit the candidate
 
-- The maintainer reviews compatibility and downstream migration impact against
-  `docs/config.md`, Go API comments, and the consumer guide.
-- The maintainer inspects `git status --short`, `git diff`, and the candidate file
-  list for credentials, private data, generated output, unintended files, and
-  unrelated changes.
-- The maintainer records a successful native checker run on Linux, including
-  filesystem capability tests. Missing required Linux evidence blocks release;
-  no macOS run, sign-off, or cross-build result is required.
-- The maintainer removes completed, obsolete documents under `docs/roadmap/`, repairs
-  inbound links, and reruns the checker. ADR rationale remains historical.
-- The maintainer reviews the exact version-matched release note for accuracy and
-  confirms it describes only deliverables that will actually be published.
+Choose an unused stable version locally and on the authoritative remote. Compare
+the candidate with the previous tag and review exported Go signatures and
+ownership, YAML and resolution semantics, diagnostic contracts, minimum Go
+version, and platform support. Review maintained consumer examples and the
+migration guidance. Explain any consumer-minimum increase and required consumer
+action under the
+[consumer minimum policy](policy/release.md#consumer-minimum-and-release-toolchain).
 
-## 5. Commit, push, and record the candidate
+Create `docs/releases/<version>.md`. Its first line must be exactly
+`# ConfigKit <version>`, and it must end with the license and attribution footer:
 
-Commit all candidate source, contract, example, roadmap-retirement, and release-note
-changes, rerun the checker against that clean commit, then push only `main` and record
-the exact commit in Git metadata local to the clone:
-
-```sh
-git add .
-git commit --message "Prepare ConfigKit $release_version"
-./scripts/release-check.sh
-test -z "$(git status --porcelain=v1 --untracked-files=all)"
-git push origin HEAD:refs/heads/main
-git rev-parse HEAD | tee .git/configkit-release-candidate
-candidate_commit=$(cat .git/configkit-release-candidate)
+```markdown
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/VERSION/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/VERSION/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/VERSION/docs/oss/attribution.md)
 ```
 
-If review changes the source after this point, create and validate a new candidate;
-do not reuse the recorded commit.
+Replace `VERSION` with the full tag and confirm the three target files exist in
+the candidate. Retire completed roadmaps, preserve unfinished and deferred
+work, and repair affected links. Review the complete candidate for unintended
+files and credentials, then commit it on `main`. These are ordinary preparation
+tasks, not separate approval gates. Any `go.mod` change belongs in this
+candidate.
 
-## 6. Run publication guards
+## 3. Push the candidate to `main`
 
-Immediately before tagging, fetch the authoritative branch and tags, restore the
-recorded variables if this is a new shell, and execute every guard:
+Push only that committed candidate to the literal authoritative URL. This does
+not follow local tags:
 
 ```sh
-release_version=v0.1.0
-candidate_commit=$(cat .git/configkit-release-candidate)
-git fetch --prune origin main
-git fetch --tags origin
-test "$(git branch --show-current)" = main
-test -z "$(git status --porcelain=v1 --untracked-files=all)"
-test "$(git rev-parse HEAD)" = "$candidate_commit"
-test "$(git rev-parse refs/remotes/origin/main)" = "$candidate_commit"
-test "$(git remote get-url origin)" = 'https://gitea.maximumdirect.net/eric/ConfigKit.git'
-printf '%s\n' "$release_version" | grep -Eq '^v(0|[1-9][0-9]*)\.(0|[1-9][0-9]*)\.(0|[1-9][0-9]*)$'
-test -f "docs/releases/$release_version.md"
-test "$(sed -n '1p' "docs/releases/$release_version.md")" = "# ConfigKit $release_version"
-for heading in Summary Compatibility Upgrade Changes; do
-  grep -Fxq "## $heading" "docs/releases/$release_version.md"
-done
-git show-ref --verify --quiet "refs/tags/$release_version" && exit 1 || :
-if git ls-remote --exit-code --tags origin "refs/tags/$release_version" >/dev/null; then
-  printf 'remote tag already exists: %s\n' "$release_version" >&2
-  exit 1
-else
-  test $? -eq 2
-fi
+git push --no-follow-tags https://gitea.maximumdirect.net/eric/ConfigKit.git HEAD:refs/heads/main
 ```
 
-Confirm the recorded automated and manual approvals apply to exactly
-`$candidate_commit`. A successful older run is not evidence for a changed commit.
+ReleaseKit requires authoritative `main` to equal this candidate before it
+validates. A failure before this push means ReleaseKit publication has not
+started. If the push result is uncertain, inspect authoritative `main` and
+continue only when it is confirmed at the candidate. If validation later fails,
+no release tag is published; correct the issue in a new commit, push the new
+candidate to `main`, and invoke validation again.
 
-## 7. Tag and publish the exact ref
+## 4. Publish the committed candidate
 
-Create one annotated tag at the guarded commit, verify its peeled target, and push
-only that ref. Do not use `git push --tags`.
+With the authorized candidate already on authoritative `main`, invoke the
+module-selected native executable exactly once:
 
 ```sh
-git tag --annotate "$release_version" --message "ConfigKit $release_version" "$candidate_commit"
-test "$(git rev-list -n 1 "$release_version")" = "$candidate_commit"
-git push origin "refs/tags/$release_version:refs/tags/$release_version"
+"$releasekit_bin" publish vMAJOR.MINOR.PATCH
 ```
 
-The pushed tag completes the declared source-only publication. No Gitea release
-object or binary upload is required.
-
-## 8. Verify publication and clean consumer retrieval
-
-Verify the authoritative tag and then retrieve that exact version into a new module
-without a local `replace` directive. Private-host credentials must already be
-configured. Use direct origin retrieval unless an explicitly approved internal Go
-proxy is the publication channel.
+Replace the placeholder with the selected version. ReleaseKit checks the
+effective push URL against `expected_push_url`, verifies the candidate on
+`main`, and checks the release-note H1 for the product and version. It then
+runs the selected race-enabled standard suite once with test-time vet disabled,
+one full vet pass, and CGO-disabled package compilation for Linux amd64 and
+arm64. It publishes only the annotated tag for that candidate. Confirmed
+publication of the intended tag completes the release. The tag points to the
+captured commit and uses the complete checked-in note as its annotation message.
+The optional `check`
+command is available for diagnosis but is not a separate required pass:
 
 ```sh
-remote_tag_commit=$(git ls-remote origin "refs/tags/$release_version^{}" | awk '{print $1}')
-test "$remote_tag_commit" = "$candidate_commit"
+"$releasekit_bin" check vMAJOR.MINOR.PATCH
+```
+
+Status 0 confirms publication, status 1 reports a known publication failure or
+conflict, status 2 reports usage, profile, or version errors, and status 3 means
+the push outcome remains unknown. Preserve both the native exit status and its
+diagnostics.
+
+The command does not push or repair `main`, wait for CI, or publish binaries or
+a hosted release page. A successful intended-tag publication does not require
+remote read-back or consumer retrieval; those checks remain optional.
+
+## 5. Recover from a failed or uncertain tag push
+
+The executable reports the publication state, phase, version, candidate commit,
+tag-object ID, effective endpoint, and error on standard error. It performs one
+bounded authoritative reconciliation after a failed tag push. Exit status 3
+means the tag outcome remains unknown; do not blindly retry or move a tag.
+
+For an uncertain result, use the exact endpoint, candidate commit, version, and
+tag-object ID from the report. Copy the captured endpoint into the variable
+below using normal shell quoting. Check that Git still resolves it to itself,
+then query the authoritative refs once:
+
+```sh
+authoritative='COPY_CAPTURED_ENDPOINT'
+release_version='vMAJOR.MINOR.PATCH'
+test "$(git ls-remote --get-url -- "$authoritative")" = "$authoritative" &&
+  git ls-remote --refs -- "$authoritative" refs/heads/main "refs/tags/$release_version"
+```
+
+A failed URL-equality check or remote query leaves the outcome unresolved.
+Correct transport configuration before continuing; a rewritten destination
+cannot establish absence at the captured endpoint.
+
+Compare the `main` object ID with the reported candidate commit and the tag
+object ID with the reported tag-object ID. `--refs` returns the annotated tag
+object itself, not merely its peeled commit. If the tag object matches, the
+release is published. If the query fails, the result is conflicting, or the
+endpoint or candidate differs, stop and resolve the state before retrying. If
+the endpoint is reachable, the tag is confirmed absent, and `main` still equals
+the reported candidate, remove a retained local tag only after confirming it
+was not published or used by another channel; then rerun `publish`, which
+revalidates the candidate:
+
+```sh
+git tag -d "$release_version"
+"$releasekit_bin" publish "$release_version"
+```
+
+A validation failure after the candidate reached `main` leaves an untagged
+candidate there. Correct it with a new commit rather than rewriting or rolling
+back `main`. A conflicting published tag must never be moved or reused. A
+defect in published source requires a corrected commit, a new semantically
+appropriate version, and a matching release note.
+
+## 6. Optional consumer verification
+
+When useful or requested, read back the remote tag or retrieve the exact
+version into a new module without a local `replace` directive. These checks are
+optional and non-gating. Do not poll CI or wait for module propagation. For
+direct-origin verification, use the exact published version in a temporary
+consumer module:
+
+```sh
+release_version=vMAJOR.MINOR.PATCH
 consumer_dir=$(mktemp -d)
 (
   cd "$consumer_dir"
   go mod init example.invalid/configkit-release-check
   GOPRIVATE=gitea.maximumdirect.net GOPROXY=direct go get "gitea.maximumdirect.net/eric/ConfigKit@$release_version"
-  cat > main.go <<'EOF'
-package main
-
-import configkit "gitea.maximumdirect.net/eric/ConfigKit"
-
-func main() {
-        _ = configkit.NoSource()
-}
-EOF
-  gofmt -w main.go
-  go test ./...
-  go build ./...
   go list -m -f '{{.Path}} {{.Version}}' gitea.maximumdirect.net/eric/ConfigKit
 )
-rm -rf -- "$consumer_dir"
 ```
 
-The final `go list` output must show the exact module path and `$release_version`.
-Record the result with the release evidence.
-
-## 9. Handle failures without replacement
-
-Before publication, fix the candidate, rerun every gate, commit, push, and record a
-new exact candidate. An unpublished local tag may be deleted only after verifying it
-was never pushed or used by another release channel.
-
-After publication, never move, delete, or recreate the tag and never replace its
-release note. Correct the defect on `main`, choose a new semantically appropriate
-version, and repeat this procedure. A supported advisory or deprecation mechanism
-may warn consumers about the old version but cannot change its contents.
+The output should show the exact module path and selected version. Report this
+optional check separately from publication success.

docs/releases/v0.4.0.md

@@ -0,0 +1,56 @@
+# ConfigKit v0.4.0
+
+This release renames selected configuration profiles to overlays, intentionally
+breaking the previous naming contract without compatibility aliases or fallback.
+
+## Summary
+
+Selected configuration now uses `Options.Overlay` and `overlays/<name>.yml`.
+Selection, merge rules, input precedence, and lifecycle behavior are unchanged.
+See the [selected-overlay contract](../config.md#selected-overlays).
+
+## Compatibility
+
+This is a breaking minor release before v1.0.0. Four surfaces change:
+
+- Go API: `Options.Profile` becomes `Options.Overlay`, `LayerProfile` becomes
+  `LayerOverlay`, and all seven `CodeProfile*` constants become `CodeOverlay*`.
+- Selected files: `profiles/<name>.yml` becomes `overlays/<name>.yml`; there is
+  no legacy-directory fallback.
+- Provenance: the layer string changes from `profile` to `overlay`.
+- Structured diagnostics: the `profile.*` code prefix becomes `overlay.*`.
+
+The old Go identifiers are removed. Application-owned concepts such as
+`ModelProfile` and `model_profile` are unchanged. Ordinary explicit imports from
+a directory named `profiles` remain valid base inputs.
+
+The minimum Go version remains 1.25. Linux remains release-validated and macOS
+remains best-effort. The YAML dependency remains pinned to
+`go.yaml.in/yaml/v4 v4.0.0-rc.6`; this release does not change parser dependencies.
+
+## Upgrade
+
+Update the dependency:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.4.0
+```
+
+Rename the affected Go identifiers, move selected files into `overlays/`, update
+any handling of provenance-layer and diagnostic-code strings, and rebuild.
+The maintained example helper is now `LoadNarratioOverlay` rather than
+`LoadNarratioProfile`. Application CLI flag names remain application-owned;
+ConfigKit does not require renaming domain-specific profile concepts.
+See the [migration guide](../consumers/README.md#selected-overlay-migration).
+
+## Changes
+
+- Rename the selected feature consistently across the public facade, internal
+  engine, diagnostics, source labels, examples, and current documentation.
+- Add regression coverage for the new directory, absence of legacy fallback,
+  public values, explanations, and preservation of application profile fields.
+- Align contributor policies with the current shared standard.
+- Simplify release validation to one local minimum-toolchain check. CI is not a
+  release gate; successful publication of the intended tag completes the
+  source-only release. See the [release procedure](../release.md).
+- Retire completed overlay and release-workflow planning documents to Git history.

docs/releases/v0.4.1.md

@@ -0,0 +1,42 @@
+# ConfigKit v0.4.1
+
+This release formalizes ConfigKit's licensing and upstream attribution
+documentation without changing library behavior.
+
+## Summary
+
+ConfigKit now includes an explicit LGPL-3.0-or-later grant in COPYRIGHT,
+complete LGPLv3 and incorporated GPLv3 license texts, and centralized upstream
+attribution. README and release guidance provide navigation to those materials.
+
+## Compatibility
+
+No Go code, public API, configuration semantics, diagnostic behavior, or
+dependencies changed from v0.4.0. The minimum Go version remains 1.25, Linux
+remains release-validated, and macOS remains best-effort. The YAML dependency
+remains `go.yaml.in/yaml/v4 v4.0.0-rc.6`.
+
+The project grant is explicitly LGPL-3.0-or-later. Third-party material remains
+subject to its respective terms, recorded in the attribution document below.
+Published earlier releases are unchanged.
+
+## Upgrade
+
+Update the module dependency and rebuild to select this release:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.4.1
+```
+
+No application-code or configuration migration is required from v0.4.0.
+
+## Changes
+
+- Add the project COPYRIGHT statement and complete standard project license texts.
+- Document go-yaml attribution with exact upstream revision links and local
+  Apache-2.0 and MIT license texts.
+- Adopt the house OSS policy and connect licensing ownership and contributor guidance.
+- Add license and attribution navigation to README and future release documents,
+  without adding release gates or changing publication mechanics.
+
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.4.1/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.4.1/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.4.1/docs/oss/attribution.md)

docs/releases/v0.4.2.md

@@ -0,0 +1,45 @@
+# ConfigKit v0.4.2
+
+This patch release simplifies and safeguards source publication and aligns
+maintainer policy and licensing guidance. Library behavior is unchanged.
+
+## Summary
+
+Maintainers can now publish a prepared release with one guarded command. It
+validates the exact committed candidate, checks the authoritative destination,
+and pushes only the candidate on `main` and its annotated version tag. The
+[release procedure](../release.md) explains preparation and partial-publication
+recovery; successful tag publication completes the source-only release without
+waiting for CI or consumer retrieval.
+
+## Compatibility
+
+No Go code, public API, configuration semantics, diagnostic behavior, or
+dependencies changed from v0.4.1. The minimum Go version remains 1.25. Linux
+amd64 and arm64 remain release-validated; macOS remains best-effort. The YAML
+dependency remains `go.yaml.in/yaml/v4 v4.0.0-rc.6`.
+
+## Upgrade
+
+Select this version and rebuild:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.4.2
+```
+
+No application-code or configuration migration is required from v0.4.1.
+
+## Changes
+
+- Add guarded single-command publication with exact candidate, version, and
+  remote checks; note-derived annotated tags; and bounded failed-push recovery.
+- Keep one offline race suite, one full vet pass, and Linux amd64/arm64 builds
+  as release gates. Formatting and whitespace remain development checks;
+  dependency tidiness is reviewed when dependencies change.
+- Align adopted policies and contributor guidance with their pinned upstream
+  revision, and include the CC BY-SA 4.0 license text and attribution for
+  adapted policy material. The project grant remains LGPL-3.0-or-later.
+- Retire completed release-workflow and policy-alignment planning records while
+  preserving their history in Git. Deferred scope remains documented.
+
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.4.2/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.4.2/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.4.2/docs/oss/attribution.md)

docs/releases/v0.5.0.md

@@ -0,0 +1,53 @@
+# ConfigKit v0.5.0
+
+This minor release adds source-backed layered configuration and built-in sibling
+inheritance, so applications can load one complete configuration while retaining
+presence, source context, and explainable derivation.
+
+## Summary
+
+- `LoadLayered` combines ordered underlay groups beneath the main input, selected
+  overlay, and explicit CLI overrides. Inputs within a group still compose
+  strictly; precedence tiers combine mappings recursively. The main input
+  remains the owner of the optional schema-version check.
+- `NewFilesystem(...).File(...)` accepts caller-owned `fs.FS` sources, and
+  `Mapping(...)` snapshots supported built-in values as a labeled source.
+  `Result.SourceContext` and `Resolution.SourceContext` expose passive resource
+  contexts for values from file-backed sources.
+- `extends` now resolves a mapping against one named sibling after source
+  precedence and before typed decoding. Inherited values count as explicit
+  input. Origins, history, explanations, and redacted rendering retain the
+  source and the structural path by which a value reached its destination.
+  Review corrections preserve immutable nested-history evidence and bounded
+  expansion work.
+
+See the [configuration contract](../config.md#layered-loading),
+[inheritance rules](../config.md#sibling-inheritance), and
+[consumer examples](../consumers/README.md) for current behavior.
+
+## Compatibility and upgrade
+
+This is an intentional pre-v1 incompatible minor release for configurations
+that use `extends` as ordinary data. That key is now reserved wherever an input
+mapping occurs; it is consumed before decoding. Rename any such YAML key and
+corresponding Go field, then rebuild. No collection registration or opt-in flag
+is available. Configurations without `extends` retain their existing meaning.
+
+Applications adopting the new features may move source-backed defaults into
+explicit `Underlay` groups, use `SourceContext` for relative resources, or
+replace application-owned sibling merging with one `extends` relationship per
+mapping. Adoption is optional. The maintained
+[named-preset example](../consumers/README.md#named-sibling-inheritance)
+shows inherited `false` and destination-local nested resolution.
+
+Select the new module version:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.5.0
+```
+
+The minimum Go line remains 1.25. Linux amd64 and arm64 remain the
+release-validated targets; macOS remains best-effort. The YAML dependency is
+unchanged at `go.yaml.in/yaml/v4 v4.0.0-rc.6`.
+
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.5.0/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.5.0/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.5.0/docs/oss/attribution.md)

docs/releases/v0.6.0.md

@@ -0,0 +1,63 @@
+# ConfigKit v0.6.0
+
+This minor release reduces avoidable configuration-loading and inspection work,
+adopts ReleaseKit for source publication, and raises the consumer Go minimum to
+1.26.0. The public configuration model remains unchanged from v0.5.0.
+
+## Changes
+
+- Reduce temporary path allocations by sizing pointer-formatting buffers,
+  combining final metadata traversal, reusing parsed sensitivity paths, and
+  avoiding path construction when detecting inheritance. Effective and
+  historical metadata indexes retain independently owned paths.
+- Replace repeated resolution-coverage, metadata-update, CLI-overlap, and
+  sensitivity-propagation scans with private indexes and incremental processing.
+  Preserve diagnostic ordering, history, contextual path shapes, and redaction.
+- Make presence-only metadata queries avoid materializing provenance. Prepare
+  resolution replay descriptors and immutable mapping and inheritance-history
+  lookups for inspection, and use ordered membership indexes for provenance
+  collections. Returned records remain detached; explanations still replay
+  complete trees and render on demand.
+- Add a maintained public-API startup construction benchmark. The initial path
+  formatting and metadata traversal experiments reproduced allocation reductions
+  across local and downstream workloads. Later algorithmic improvements were
+  validated behaviorally without a new benchmark campaign. Busy-host timings
+  were inconclusive; no fixed speedup or allocation threshold is promised.
+- Replace ConfigKit-owned release scripts with a small profile and the
+  module-selected ReleaseKit v0.3.0 tool. The
+  [release procedure](../release.md) documents candidate preparation, tag-only
+  publication, and failed-push recovery. Existing release checks and the
+  successful-tag-push completion point remain unchanged; CI is not a gate.
+- Adopt release policy 1.6, distinguishing the consumer Go minimum from the
+  validation toolchain. Retire completed planning and measurement records into
+  Git history while preserving [performance candidates](../roadmap/future.md#3-performance-experiments).
+
+## Compatibility and upgrade
+
+The consumer minimum increases from Go 1.25.0 to **Go 1.26.0**. Install Go 1.26.0
+or newer before upgrading, and update application build and CI environments that
+still use Go 1.25. The `go.mod` directive records the consumer minimum; using a
+newer validation patch does not itself raise that minimum. See the
+[consumer minimum policy](../policy/release.md#consumer-minimum-and-release-toolchain).
+
+Select this version and rebuild:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.6.0
+```
+
+No application-code or configuration migration is required from v0.5.0. Public
+Go signatures, strict within-tier composition, recursive tier precedence,
+sibling inheritance, presence/null distinctions, provenance, secret handling,
+and synchronous completion retain their existing contracts. See the
+[configuration contract](../config.md) and
+[consumer guide](../consumers/README.md).
+
+The YAML dependency remains `go.yaml.in/yaml/v4 v4.0.0-rc.6`. ReleaseKit is a
+declared Go tool dependency for maintainer commands, not a library runtime
+import; its requirements participate in ConfigKit's module graph. Linux amd64
+and arm64 remain release-validated targets, and macOS remains best-effort.
+Maintainers should use the current release procedure instead of the retired
+`scripts/release.sh` and `scripts/release-check.sh` commands.
+
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.6.0/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.6.0/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.6.0/docs/oss/attribution.md)

docs/releases/v0.7.0.md

@@ -0,0 +1,58 @@
+# ConfigKit v0.7.0
+
+This minor release adds structured comparison of completed configurations.
+Applications can inspect changes between two configuration files or the
+effective impact of a selected overlay, with ConfigKit handling comparison and
+safe value disclosure while applications own presentation.
+
+## Changes
+
+- Add `Diff[T](left, right *Result[T]) ([]Difference, error)` through the public
+  facade. It compares frozen effective inspection values from successful loads,
+  including defaults and recorded application resolution, rather than source
+  text or later mutations of `Result.Config`.
+- Return detached changed-region records containing a semantic path, safe left
+  and right `ValueSnapshot` values, and `HiddenChanged`. Mappings compare
+  recursively, sequences report atomic changes, and records follow canonical
+  pointer order. Presence distinguishes additions and removals from explicit
+  null; equal effective values produce no records regardless of provenance or
+  loading history.
+- Protect both snapshots with sensitivity from both loads, including historical
+  and dependency-derived classification. Compatible regions retain public
+  siblings; incompatible structural replacements use complete-region masking
+  where required. `HiddenChanged` identifies actual changes withheld by masking,
+  rather than merely indicating that a snapshot contains redacted content.
+- Preserve visible public scalar changes when only historical sensitive
+  descendants remain, including additions and removals. Historical explanation
+  redaction and masking of incompatible present shapes remain intact.
+- Add maintained file-comparison and base-versus-overlay examples with
+  consumer-owned formatting, plus an executable public API example. See the
+  [consumer guide](../consumers/README.md#effective-configuration-diffs) and
+  [configuration contract](../config.md#effective-configuration-diffs).
+
+## Compatibility and upgrade
+
+No application-code or configuration migration is required from v0.6.0.
+Adopting the new API is optional. Loading, composition, inheritance, provenance,
+existing inspection, and secret-handling contracts remain unchanged.
+
+Load both configurations successfully before calling `Diff`; nil or
+uninitialized results return an error with no records. Comparison does not
+reopen sources, rerun callbacks, or reread environment variables. Typed secret
+references compare by configured reference name, not resolved credential bytes,
+so credential rotation behind an unchanged name is not reported. Returned
+snapshots describe changed regions, not cross-protected full-document context.
+
+Select this version and rebuild:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.7.0
+```
+
+The consumer minimum remains Go 1.26.0. The YAML dependency remains
+`go.yaml.in/yaml/v4 v4.0.0-rc.6`, and the maintainer tool dependency remains
+ReleaseKit v0.3.0. Linux amd64 and arm64 remain release-validated targets;
+macOS remains best-effort. Publication follows the existing
+[release procedure](../release.md).
+
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.7.0/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.7.0/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.7.0/docs/oss/attribution.md)

docs/releases/v0.7.1.md

@@ -0,0 +1,22 @@
+# ConfigKit v0.7.1
+
+This patch updates the maintainer release tool to ReleaseKit v0.4.1. ConfigKit's
+runtime behavior and public API are unchanged.
+
+## Compatibility and upgrade
+
+No application-code or configuration migration is required from v0.7.0. The
+consumer minimum remains Go 1.26.0, and the YAML dependency is unchanged.
+
+Select this version and rebuild:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.7.1
+```
+
+ReleaseKit v0.4.1 is a maintainer tool dependency and does not change consumer
+configuration or runtime behavior. Linux amd64 and arm64 remain
+release-validated targets; macOS remains best-effort. Publication follows the
+existing [release procedure](../release.md).
+
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.7.1/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.7.1/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.7.1/docs/oss/attribution.md)

docs/releases/v0.8.0.md

@@ -0,0 +1,70 @@
+# ConfigKit v0.8.0
+
+This minor release adds explicit input placement, relative import destinations,
+shared source-version checking, and opt-in composition of identified sequence
+records. Applications can assemble independently authored configuration fragments
+into one final configuration while retaining source locations, provenance,
+history, and sensitivity information.
+
+## Changes
+
+- Add `Destination` to ordinary load options and layered inputs. A source mapping
+  can contribute beneath literal application keys without being rewritten by the
+  consumer. Imports inherit that destination or append a relative YAML
+  `destination`. Physical import paths and resource contexts remain relative to
+  their declaring sources. See [input destinations](../config.md#input-destinations),
+  [import syntax](../config.md#explicit-imports), and the
+  [mounted collection example](../consumers/README.md#input-destinations-and-mounted-collections).
+- Add `SourceVersionModeIfPresent` and `SourceVersionModeRequired` to apply one
+  application-owned version checker across main, underlay, imported, and selected
+  overlay documents. Enabled modes check source-local declarations and consume
+  accepted non-main headers before composition. Detached, sensitivity-aware
+  evidence is available through `SourceInfo.VersionCheck`. Acceptance remains
+  application policy; ConfigKit does not require declarations to equal the main
+  version. See [shared source-version checks](../config.md#shared-source-version-checks).
+- Add explicit `PlacementModeSequenceItem` placement for ordinary and layered
+  inputs, and `placement: sequence-item` for path and glob imports. A source
+  mapping becomes one item at its destination; ordinary helper imports compose
+  within that item and can supply its identity. See
+  [imports](../config.md#explicit-imports).
+- Add `Definition.SequenceCompositions` registrations for strict composition at
+  selected literal paths using one string identity field. Distinct records append
+  in contribution order, while duplicate identities conflict. Higher-priority
+  sequences still replace lower-priority sequences wholesale. Registration does
+  not add identity-based overlays or final identity validation after defaults and
+  application resolution. See the [composition contract](../config.md#strict-document-composition)
+  and [maintained prompt catalog](../consumers/README.md#identified-prompt-sequences).
+- Extend source inspection with destination and placement evidence, including
+  whether a nested destination is unresolved in an intermediate or failed-load
+  snapshot. Preserve source-specific history, consumed-header sensitivity, and
+  explanations of descendants removed by sequence replacement as records are
+  assembled. Source locations continue to identify authored coordinates.
+
+## Compatibility and upgrade
+
+The new behavior is opt-in: zero-valued destinations retain root placement,
+mapping placement remains the default, unregistered sequences retain strict
+conflict behavior, and the default source-version mode preserves the existing
+main-only version policy. Recursive mapping precedence, whole-sequence
+replacement between tiers, and application-owned final validation remain intact.
+
+Public option, definition, input, and source-information structs gain fields.
+Consumers using unkeyed literals for affected structs must switch to keyed
+literals or update them. `SourceInfo` now contains a `Path` and is no longer
+comparable with `==` or usable directly as a map key; compare the relevant fields
+or select an application-appropriate key instead. Existing keyed literals need
+no changes to retain the previous loading behavior.
+
+Select this version and rebuild:
+
+```sh
+go get gitea.maximumdirect.net/eric/ConfigKit@v0.8.0
+```
+
+The consumer minimum remains Go 1.26.0. The YAML dependency remains
+`go.yaml.in/yaml/v4 v4.0.0-rc.6`, and the maintainer tool dependency remains
+ReleaseKit v0.4.1. Linux amd64 and arm64 remain release-validated targets;
+macOS remains best-effort. Publication follows the existing
+[release procedure](../release.md).
+
+[COPYRIGHT](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.8.0/COPYRIGHT) | [LICENSE](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.8.0/LICENSE) | [ATTRIBUTION](https://gitea.maximumdirect.net/eric/ConfigKit/src/tag/v0.8.0/docs/oss/attribution.md)

docs/roadmap/deferred.md

@@ -1,28 +0,0 @@
-# Deferred Scope
-
-This document owns intentionally excluded features, not a committed delivery plan.
-The [architecture policy](../policy/architecture.md) governs changes to these
-boundaries. Implemented behavior belongs in the [configuration contract](../config.md)
-and consumer adoption in the [consumer guide](../consumers/README.md).
-
-## Explicit non-goals and deferred scope
-
-General environment-variable configuration overlays are not supported. Typed secret
-references and explicit directory bootstrap are the environment convention. Literal
-credential configuration and implicit secret-directory discovery are unsupported.
-
-The initial release deliberately excludes:
-
-- automatic `conf.d` discovery;
-- remote imports, arbitrary providers, and additional configuration formats;
-- custom merge strategies, sequence merging, reset operators, and profile inheritance;
-- parser-specific or custom structural-decoding extension APIs;
-- generic dynamic-subtree conversion and module-registry APIs;
-- automatic schema migrations;
-- hot reload, runtime resource replacement, and remote resource validation;
-- a bundled CLI framework or application commands; and
-- configuration diffs and fingerprints.
-
-Applications may build their own validate, effective-value, source, bootstrap, and
-explanation commands around the library. Broadening these boundaries requires
-concrete consumer evidence, compatibility review, and an explicit design change.

docs/roadmap/future.md

@@ -0,0 +1,178 @@
+# Potential Future Features
+
+This document lists potential enhancements for consideration in future ConfigKit
+releases. Entries are not accepted contracts, implementation plans, or release
+commitments. Adopting a candidate requires an explicit design decision and a
+separate plan. Significant architectural exclusions and their rationale belong
+in ADRs; this document does not maintain a list of rejected proposals.
+
+## 1. Identified record sequences
+
+Status: deferred. PromptKit proposed this capability from its retired proposal 2,
+**Namespaced Collections As Identified Records**. The original proposal is in the
+PromptKit repository at
+`b5866dd:docs/roadmap/configuration-collections-identified-records.md`.
+PromptKit selected mounted mappings instead; this feature is not a dependency
+of that design or a request to restore multiple installed prompt versions.
+
+A narrower [sequence placement and composition feature](../config.md#strict-document-composition)
+is implemented: consumers can place imported mappings as sequence items and
+register a single literal string identity field for duplicate-rejecting strict
+composition. Precedence still replaces sequences as whole values. This entry
+tracks broader identified-record behavior that remains deferred and is not part
+of that implemented contract.
+
+The active [sequence roadmap](sequences.md) now selects a narrower design for
+collection policy, compound scalar identities, final application-registered
+identity validation, and inheritance within identified sequences. It owns those
+selected behaviors and their remaining decisions. The original broader proposal
+below remains background for further consideration, particularly recursive
+combination of matching record bodies; it does not override the active roadmap.
+
+### Motivation and proposed behavior
+
+Allow applications to declare that selected sequences contain records identified
+by explicit fields, while retaining sequence-shaped output for slice decoding
+and effective YAML. This could make composite identities and independently
+authored catalogs more convenient than encoding identity in nested mapping keys.
+Current [composition](../config.md#strict-document-composition) appends distinct
+records only at explicitly registered paths; other sequences remain atomic.
+Selected [overlays](../config.md#selected-overlays) always treat sequences
+atomically.
+
+The proposed broader opt-in registration would extend the current reusable
+application definition from one identity field to a composite identity made of
+an ordered, nonempty set of distinct direct string-field names. For example,
+records at `profiles` could be identified by `(namespace, id)`. Records
+`(provider, brief)`
+and `(application, brief)` would coexist, and a later layer could update the
+provider record's temperature while retaining its backend and model.
+
+The requested semantics, all still subject to design review, are:
+
+- Match exact string tuples, preserving component boundaries and spelling,
+  including empty strings. Every contribution supplies its complete identity;
+  defaults, inheritance, filenames, and normalization do not supply it.
+- Retain distinct identities. Matching records from strict peers recursively
+  compose disjoint body fields and conflict on overlapping non-mapping body
+  values, even equal values. Equal identity fields act as structural keys rather
+  than conflicts and retain their contributing source evidence.
+- Higher layers recursively overlay matching bodies. A changed identity denotes
+  a different record, not a rename. Proposed ordering is first contribution in
+  precedence-assembly order; matching records keep position and new ones append.
+- Reject malformed identities, non-mapping records, non-null non-sequence values
+  at registered paths, and duplicate tuples within one source sequence. Later
+  layers cannot repair malformed structural identity in an acquired contribution.
+- Treat an empty registered sequence as contributing no records; whole-path null
+  retains ordinary conflict/replacement semantics. Unregistered sequences,
+  including nested message lists, retain existing atomic behavior.
+- Support record inheritance within the same registered sequence using an
+  `extends` mapping containing the complete parent tuple. Resolve only final
+  active relationships after precedence, preserve the child's identity, overlay
+  its body on the parent's body, and consume the directive before decoding.
+  Preserve cancellation, missing-parent/cycle errors, and declaring resource
+  contexts. Existing string-based mapping-sibling inheritance stays unchanged.
+
+Integrate ordinary and layered loads, imports, mounted paths, selected overlays,
+and whole-sequence CLI overrides. Keep origins, presence, history, sensitivity,
+and resource contexts attached to the correct record despite different source
+and final indexes. Diagnostics must retain source locations without exposing
+sensitive identity values. Final inspection may use snapshot-specific sequence
+indexes; existing atomic sequence diffs need not become identity-aware.
+
+Successful results would have valid, unique identities, including after defaults
+and application resolution, without a second merge or inheritance pass. Existing
+source-version checks remain independent; consumed headers cannot supply missing
+record identities. No identity-selector path syntax, arbitrary merge callbacks,
+wildcard registration, per-record deletion/rename operator, multiple inheritance,
+or cross-path parent references are requested.
+
+### Questions to resolve before implementation
+
+- **Need and delivery scope:** does an active consumer benefit enough to justify
+  the added semantics over keyed mappings? Are composite identities required?
+  Staging composition before inheritance is possible but would leave part of
+  the request incomplete.
+- **Defaults and callbacks:** do default records merge by identity or retain
+  existing sequence-replacement behavior? How are callback-created duplicates,
+  identity edits, and reordering validated and reflected in provenance?
+- **Null, empty, and order:** what happens when null clears a collection and a
+  later layer reintroduces records? How do absent and explicitly empty sequences
+  interact with defaults and first-contribution order?
+- **Inheritance references:** how do peers compose relationship declarations,
+  and how are complete references enforced without assembling partial tuples
+  accidentally? What happens if identity fields overlap structural controls?
+- **Registration paths:** how do overlapping registrations and paths beneath
+  sequence indexes behave when assembly changes those indexes?
+- **Evidence and lifecycle:** how are record identities mapped between source
+  trees, history, inheritance, callback snapshots, and final indexes without
+  mixing records' provenance or sensitivity? Where do identity checks run relative
+  to header consumption and other structural processing?
+
+Keep the feature deferred until its consumer value and full lifecycle design
+justify reconsidering the sequence-merging exclusion. A correct final sequence
+alone would not establish correct inspection or disclosure behavior.
+
+## 2. Required version declarations for selected sources
+
+Status: deferred. PromptKit requested a way to require version declarations on
+a selected subset of source files while permitting other fragments to omit them.
+
+### Motivation and proposed behavior
+
+PromptKit could require each collection's root file (for example,
+`collection.yml` or `meta.yml`, with naming still undecided) to declare an
+accepted specification version. Other files would use the existing
+[check-when-present behavior](../config.md#shared-source-version-checks).
+This would enforce a rule that otherwise remains a documentation convention.
+
+All declarations would use the same application-owned `Definition.Version`
+path and acceptance policy. The proposal concerns where declarations are
+required; it does not introduce independently versioned collections or subtrees.
+A root declaration would not prove that headerless descendants declare a
+compatible version or replace final configuration validation.
+
+### Questions to resolve before implementation
+
+- **Selection API:** should callers mark explicit inputs, select sources by
+  path or predicate, or annotate imports? An input-level requirement is simpler,
+  but cannot by itself select a root reached only through an import.
+- **Scope:** does a requirement apply only to the selected document or also to
+  its imports? Applying it only to that document best fits PromptKit's request;
+  imported fragments could still declare versions checked by the shared policy.
+- **Mode interaction:** how should selection interact with disabled source
+  checks? Require an enabled mode, or activate checking for selected sources?
+  Selection would strengthen check-when-present behavior and add no requirement
+  beyond the load-wide required mode.
+- **Source handling and evidence:** how should selection work across file,
+  reader, byte, and mapping inputs, glob matches, and repeated imports? Define
+  behavior for absent optional inputs and preserve source-local diagnostics,
+  inspection evidence, and existing sensitivity handling.
+
+## 3. Performance experiments
+
+Status: deferred pending evidence. The completed performance audit identified
+four internal optimization candidates for separate investigation:
+
+- Branch-local explanation replay instead of the current whole-tree replay.
+- Redesigning the private metadata-copy handoff while preserving detached public
+  records.
+- Caching schema descriptors across loads.
+- Indexing complex contributor-trace membership while preserving its equality
+  rules and compact linked traces.
+
+These are experiments, not committed implementation work or promised speedups.
+Reconsider them with evidence that the benefit justifies the ownership,
+retained-memory, or equivalence risks. The completed audit and implementation
+evidence are preserved in Git history at `8e91a2b` under `docs/roadmap/`.
+
+### Questions to resolve before implementation
+
+- Which representative workloads show enough remaining cost to justify each
+  optimization, and what measurements would establish a useful improvement?
+- Can narrower replay and membership indexes preserve explanation results,
+  ordering, provenance, and sensitivity behavior?
+- Can metadata handoff changes preserve detached public records and ownership
+  across loads without introducing aliasing?
+- What cache lifetime and concurrency design would make schema reuse worthwhile
+  without excessive retained memory?

docs/roadmap/implementation.md

@@ -0,0 +1,894 @@
+# Sequence Identity And Combination Implementation Plan
+
+Status: ready for implementation; no stages completed.
+
+Audience: a Luna coding agent implementing one numbered stage per prompt, in
+order. This plan starts at commit
+`4f941f8ba4c9a9ed0afa87b07fe6ba29f131d1c6`. It implements the settled
+[sequence roadmap](sequences.md), [ADR 0017](../adr/0017-collection-sequence-policy.md),
+and [ADR 0018](../adr/0018-cli-after-structural-resolution.md).
+
+## Working Rules
+
+- Read [development guidance](../development.md) and the applicable policies,
+  especially [agents](../policy/agents.md), [architecture](../policy/architecture.md),
+  [testing](../policy/testing.md), and [documentation](../policy/documentation.md).
+  Prefer the repository knowledge graph for code discovery.
+- The roadmap owns the selected feature semantics. This plan supplies execution
+  order and implementation choices, not additional product requirements. Preserve
+  unchanged source, parser, version, schema, ownership, and inspection contracts.
+- Complete each stage's implementation and behavioral tests together. Record its
+  completion and validation here, then commit the nonempty unit with a short,
+  plain-English message. Preserve unrelated changes. Do not publish intermediate
+  stages or a release, update consumers' dependencies, or wait for CI.
+- Run focused tests while iterating and `./scripts/check.sh` before handing off
+  each code stage. Use the installed compatible Go toolchain with dependencies
+  already acquired, as the development guide describes. Add race coverage at the
+  ownership milestones below. These are implementation checks, not new release
+  gates. Reuse successful checks for unchanged content rather than repeating them
+  at a documentation-only milestone. Do not require human platform validation.
+- Update affected declaration comments and canonical current-behavior prose when
+  behavior lands. Keep incomplete target behavior in this roadmap/plan until it
+  works. The final documentation stage reconciles the cross-feature narrative;
+  it is not permission to leave misleading current contracts in earlier stages.
+- Transitional internal adapters are acceptable where specified. Remove them by
+  completion. Do not add public compatibility aliases, permanent feature flags,
+  alternate semantics, new dependencies, new resource caps, or an extension/plugin
+  framework for this work.
+- Preparation stages keep existing public behavior working and test new machinery
+  privately. Activation stages integrate values, policy, metadata, and relevant
+  failure handling together; do not defer correct inspection to a later stage.
+
+## Implementation Decisions
+
+### Public Registration Surface
+
+Replace `SequenceComposition` and `Definition.SequenceCompositions` with:
+
+```go
+type SequenceIdentity struct { /* private, root-owned registration state */ }
+type SequenceSelector []SequenceSelectorSegment
+type SequenceSelectorSegment struct { /* private literal-or-any-mapping-key state */ }
+
+func SequenceLiteralKey(key string) SequenceSelectorSegment
+func SequenceAnyMappingKey() SequenceSelectorSegment
+func SequenceIdentityAt(path Path, fields ...string) SequenceIdentity
+func SequenceIdentityMatching(selector SequenceSelector, fields ...string) SequenceIdentity
+
+// Field on Definition[T]:
+// SequenceIdentities []SequenceIdentity
+```
+
+For example, the reusable definition can contain both registrations:
+
+```go
+SequenceIdentities: []configkit.SequenceIdentity{
+    configkit.SequenceIdentityAt(configkit.Keys("prompts"), "id"),
+    configkit.SequenceIdentityMatching(configkit.SequenceSelector{
+        configkit.SequenceLiteralKey("collections"),
+        configkit.SequenceAnyMappingKey(),
+        configkit.SequenceLiteralKey("prompts"),
+    }, "id"),
+},
+```
+
+The exact constructor converts mapping-only `Path` segments into literal selector
+segments. Preserve invalid index segments long enough to report an ordinary
+definition error; do not panic, silently drop them, or change `Path` semantics.
+The zero selector is the exact root selector, harmless under the existing
+mapping-root schema. A zero selector segment denotes the literal empty key,
+not a wildcard. Empty field names are valid literal field names; an empty field
+set is invalid. Constructors snapshot slices; load-entry conversion also takes
+an owned snapshot. Definition validation, not construction, reports errors.
+Root types and functions must not expose internal types or bridge methods.
+
+### Identity Representation And Matching
+
+- Normalize identity field sets once: deduplicate and sort exact field names.
+  Use the same normalizer for YAML and application registration.
+- Encode a record's identity as a hashable, type-tagged, length-delimited tuple in
+  normalized field order. Mapping lookup indexes avoid scanning record fields
+  repeatedly. Distinguish structural-payload lookup from final typed-payload and
+  parent-reference lookup: only fields acting as directives are excluded, not
+  every occurrence of the same literal spelling. Never include tuple encodings
+  in diagnostics.
+- Normalize finite JSON numbers to sign, significant decimal digits, and a signed
+  base-10 exponent. Strip insignificant zeroes and give all zero spellings one
+  representation. Use `math/big.Int` for exponent arithmetic, not a power-of-ten
+  expansion or a float conversion. Thus large exponent spellings do not allocate
+  storage proportional to their mathematical magnitude. This comparison-only
+  representation does not rewrite document nodes or typed projection.
+- Compile selectors into a mapping-segment trie with literal and wildcard edges.
+  Carry applicable trie states through traversal; stop at sequence-index edges.
+  Index concrete matches within one assembly/inheritance scope where useful, but
+  do not save an assembly-only destination list for final validation.
+- Check registration overlap pairwise once during definition validation using
+  equal length and compatible corresponding segments. This straightforward
+  `O(R² × D)` check, for R registrations and maximum selector depth D, is
+  proportionate for registration definitions. Equivalent constraints coalesce;
+  conflicting overlaps fail even with no matching input.
+  Do not impose a registration-count limit to simplify this operation.
+
+### Policy And Evidence Storage
+
+Keep application constraints separate from YAML policy. Add private engine
+collection-policy records containing normalized identity, explicit mode, and
+declaration locations. Track whether identity is pending or belongs to a retained
+collection. Retain declaration presence separately from inferred defaults.
+
+Use the existing acquisition/item scopes as ownership anchors, with an opaque
+per-load owner token when a retained record needs one. Store item-local policies
+relative to their owning record, and project concrete paths when that record is
+placed. Do not key ownership solely by numeric index, identity tuple, or source
+filename: all three can coincide for a replacement that must lose old policies.
+Inheritance clones the relevant scoped state; it does not create registrations.
+
+Resolve policies through shared engine helpers used by strict composition,
+precedence, and inheritance. Do not attach engine policy to public values or add
+policy fields to the parser-independent `document.Node`. The sidecar can be
+extended with concrete-path indexes; it need not become a general event store.
+
+Sequence combination returns the resulting list plus explicit input-to-output
+item projections and retained/replaced contribution information. Use that same
+result for values, policy transfer, origins, source destinations, sensitivity,
+and historical replay. Extend the existing sequence-history descriptors for
+noncontiguous upsert positions rather than inferring them from equal values.
+Sharing immutable internal evidence is acceptable under existing ownership rules;
+independently returned mutable records must remain detached.
+
+### Diagnostics
+
+Use existing `options.invalid` for invalid application definitions and existing
+`compose.invalid` / `compose.conflict` for malformed or duplicate strict-peer
+contributions. Add root-owned and engine diagnostic constants
+`CodeSequenceInvalid` (`sequence.invalid`) and `CodeSequenceConflict`
+(`sequence.conflict`) for malformed collection declarations/operations or final
+identities, and conflicting policy / precedence duplicates respectively.
+Final duplicate identities use `sequence.conflict`. Keep existing inheritance
+invalid/missing/cycle codes for record relationships. Use existing override and
+schema/decode codes at the CLI boundary; do not invent a new error taxonomy there.
+
+Report collection/item paths and available locations, not identity values,
+normalized tuple strings, raw parent references, or raw hook errors. Declaration
+conflicts include both sites. Definition overlap errors identify registration
+positions without fabricating a concrete matched configuration path. Exact
+English diagnostic wording is not a new guarantee.
+
+## Stage 1 — Implement Shared Scalar Identity Primitives
+
+Scope: new private identity helpers and their focused tests; no load-path change.
+
+1. Add `internal/engine/sequence_identity.go`. Implement field-set normalization,
+   scalar tuple construction, and exact decimal normalization as chosen above.
+   Accept string, number, boolean, and explicit null nodes; distinguish categories
+   and absence. Exclude fields acting as structural directives from payload lookup.
+2. Validate mapping records, complete direct-field sets, and uniqueness using one
+   tuple-to-first-record index. Return paths and source locations, not values.
+   Support caller-selected diagnostic families for strict and other boundaries.
+3. Cover equivalent field sets; empty strings and null; missing/non-scalar fields;
+   literal punctuation; compound tuple boundaries; `1`/`1.0`/`1e0`; signed zero;
+   close unequal decimals; and large exponents without expansion. Verify that
+   node text remains unchanged. Include `extends: null` as a structural field
+   that cannot supply source identity.
+
+Validation: `go test ./internal/engine -run 'SequenceIdentity|SequenceNumber'`
+and the everyday check. Retain the old registration path until its adapters are
+replaced; do not generalize unrelated YAML scalar decoding.
+
+## Stage 2 — Compile Exact And Dynamic Application Constraints
+
+Scope: private registration model, selector matching, overlap validation.
+
+1. Extend `sequence_registration.go` with the common registration representation
+   and compiled selector trie. Represent literal and arbitrary-mapping-key segments
+   structurally, not with a magic string. Snapshot registration inputs.
+2. Validate field sets and mapping-only selectors before source acquisition.
+   Preserve existing destination validation without adding restrictions on identity
+   field spelling. Root and unused selectors remain harmless.
+3. Implement full-destination matching, equivalent overlap coalescing, and
+   incompatible-overlap errors. Exact rules have no priority over dynamic rules.
+   Keep a temporary conversion from the current single-field registration so
+   existing callers still use their current behavior.
+4. Test exact and wildcard matches, multiple wildcard levels, empty/punctuation
+   keys, sequence-index nonmatches, disjoint lengths/literals, equivalent sets,
+   incompatible overlaps with absent input, and independent uniqueness scopes.
+
+Validation: focused registration/selector tests and the everyday check. This
+stage does not export the replacement API or add final-result validation yet.
+
+## Stage 3 — Parse Collection Declarations Without Filesystem Coupling
+
+Scope: private parser and directive extraction helpers, initially unit-tested
+without activating collection combination in public loads. Keep source-form
+activation out of this stage; ordinary sources and selected overlays are wired
+in Stages 13 and 14 respectively.
+
+1. Add `sequence_declaration.go` for source-root `configkit.sequences`. Each entry
+   has a required existing literal `destination` form and optional `key` and
+   `mode`. `key` is a string or nonempty string list; `mode` is exactly replace,
+   append, or upsert. Use shared identity normalization. A destination-only entry
+   can be a harmless no-op; do not add an unrequested ban on empty declarations.
+2. Split common root-directive extraction from import-specific parsing in
+   `import_spec.go` / `parse.go`. Extract imports and sequences together, retaining
+   source-local locations and explicit presence. Reject unrecognized directive
+   fields and wrong supplied field types using existing strict parsing style.
+3. Keep declaration destinations relative to source placement, with item-local
+   coordinates retained until their scope is assembled. Do not project them
+   prematurely to a guessed numeric index.
+4. Give extraction an explicit overlay context: sequences are allowed, imports
+   are not. Import enablement remains separately enforced; do not turn on
+   filesystem access because a source declares sequences. Keep enabled
+   source-version checks in raw source coordinates before consuming controls;
+   extend existing actual header/control-overlap handling to sequence directives
+   without treating an absent header beneath a control as an actual overlap.
+5. Test key-only, mode-only, combined, repeated-field-set equivalence, no-op,
+   malformed types/modes/destinations, mixed imports/sequences, and overlay rules.
+   Do not add import-level key/mode shorthand or sequence wrappers.
+
+Validation: focused declaration/parser tests, existing import-spec tests, and the
+everyday check. Existing import-only behavior remains intact during preparation.
+
+## Stage 4 — Reconcile Declarations And Select Collection Operations
+
+Scope: pure agreement and operation-selection helpers, independent of loading
+and of the ownership transformations introduced next.
+
+1. Add `collection_policy.go`. Gather the current scope's declarations before
+   combining peers. Reconcile equivalent keys and agreeing explicit modes;
+   conflicting peers fail. Omitted fields have no preference. Apply authoritative
+   application constraints even for key-only declarations without a value.
+2. Implement shared operation selection: current explicit mode, retained explicit
+   mode, inferred upsert with effective identity, otherwise replace. Persist only
+   explicit modes. Policy-only declarations must not replay earlier combinations.
+3. Accept eligible lower policy as an input; do not decide record survival or
+   ancestor removal inside declaration reconciliation. Retain declaration
+   locations/presence separately from inferred mode and application constraints.
+4. Test equivalent/different field sets, agreeing/conflicting peer modes,
+   omission, explicit default modes, and operation inference. Include key-only
+   declarations agreeing or conflicting with application constraints before a
+   sequence exists.
+
+Validation: focused policy-agreement/selection tests and the everyday check.
+Do not activate collection combination or implement the ownership layer here.
+
+## Stage 5 — Track Policy Lifetime And Record Ownership
+
+Scope: private scoped-state transformations built on Stage 4's resolution helpers.
+
+1. Implement key agreement across retained values, replacement reset, pending
+   key declarations, and immediate validation when a key-only declaration
+   identifies retained records.
+   Changing mode to replace without a replacement value does not reset identity.
+2. Implement scope/record ownership transfer and retirement hooks. Actual ancestor
+   replacement removes descendant collection identity and removed-record policy;
+   independently owned modes and application constraints survive as specified in
+   the roadmap. Do not repeatedly reapply old declarations and resurrect reset
+   identity. Pending declarations retain their owning-scope rules.
+3. Use item/acquisition scope anchors and record-owned relative policy as selected
+   above. Provide transformations for retention, whole-record replacement,
+   whole-sequence replacement, and inheritance copies without wiring them into
+   public loading yet. Application registrations remain destination constraints.
+4. Use table-driven tests derived from the roadmap's lifetime, persistence, and
+   replacement tables, including absent/non-sequence destinations, equal outer
+   identities with distinct record owners, and independent root versus item policy.
+
+Validation: focused collection-policy tests and the everyday check. This stage
+must not add path-existence, shape, nullability, or global YAML final-check rules.
+
+## Stage 6 — Implement The Shared Sequence Operations
+
+Scope: a pure document-level combination helper, not separate implementations
+for overlays and inheritance.
+
+1. Add `sequence_combine.go`, accepting lower/higher sequence nodes, resolved
+   policy, and combination context. Reuse Stage 1 identity extraction and indexes.
+2. Implement replace, duplicate-rejecting append, and whole-record upsert. Upsert
+   retains matched lower positions and appends new higher identities in order.
+   Return item projections and replacement/retention evidence with the new value.
+3. Keep peer validation distinct: duplicate identities within any contribution or
+   among peers fail under every mode. Do not merge record payloads or sort by ID.
+4. Cover absent versus null/non-sequence lower values; higher omission, `[]`, null,
+   and non-sequence values; missing identity under append/upsert; empty inputs;
+   stable ordering; equivalent/different keys; and explicit replacement identity
+   reset. An empty input does not bypass kind checks or identity requirements.
+5. Verify clones do not alias mutable input nodes or policy slices. Build indexes
+   once per operation rather than scanning the accumulated list for every record.
+
+Validation: focused combination tests and the everyday check. Do not change the
+ordinary recursive mapping rule or add unkeyed append.
+
+## Stage 7 — Extend Sequence History Representation And Replay
+
+Scope: represent and replay sequence operations without yet changing all metadata
+projection consumers or public precedence.
+
+1. Extend `sequenceHistoryContribution`, sequence replay in `history.go`, and
+   its contribution indexes to consume Stage 6's explicit item mappings. Existing
+   append ranges may remain a compact case; upsert cannot be represented by one
+   offset.
+2. Preserve authored raw contribution snapshots and displaced values. A replaced
+   record's omitted fields must appear as removed, not as retained lower payload.
+   Preserve equal-valued replacement events. Keep existing index-based query and
+   atomic sequence-diff APIs; add no public identity-selector path syntax.
+3. Exercise synthetic append/upsert/replace histories, noncontiguous matches,
+   nested paths, and discarded descendants. Compare replayed final values with
+   the actual combination result, including first/last and unchanged neighboring
+   records. Reuse existing escaped-path and detached-query tests rather than
+   duplicating them.
+
+Validation: sequence/history/replay tests and the everyday check. Keep current
+append/replace callers working through the extended descriptors.
+
+## Stage 8 — Project Sequence Origins And Inspection Evidence
+
+Scope: use the prepared item mappings consistently in metadata consumers.
+
+1. Project composition origins, nested source destinations, historical paths, and
+   sensitivity through Stage 6's item mapping and Stage 7's history descriptors.
+   Physical source IDs/locations and resource contexts continue to identify the
+   declaring source. Reuse one projection mechanism rather than searching again
+   by identity or equal payload.
+2. Test synthetic noncontiguous upserts, replaced descendants, nested source
+   destinations, unchanged neighbors, and secret-marked replaced records. Verify
+   projected queries agree with replayed and actual final values, retaining
+   displaced evidence and detached results.
+3. Keep the new precedence operations unactivated until their value, policy, and
+   projection machinery are integrated in the later load-path stages.
+
+Validation: metadata/history/sequence/provenance/sensitivity tests,
+`go test -race ./internal/engine`, and the everyday check.
+
+## Stage 9 — Separate Per-Input Acquisition From Composition
+
+Scope: a behavior-preserving acquired-input handoff; grouped underlay coordination
+is the next stage.
+
+1. In `load_input.go`, `imports.go`, and related graph helpers, retain acquisition
+   scopes, raw contributions, inclusion edges, source records,
+   contexts, and extracted declarations without irrevocably composing them.
+   Reuse existing `deferAssembly` and acquired-scope mechanisms instead of adding
+   a second loader. Keep already acquired document handles closed as before.
+2. Separate acquisition from the per-input composition entry. Retain pending
+   item/helper scope boundaries, namespace identities, inclusion/repetition checks,
+   source records, accounting, and scheduler inputs. Do not reacquire sources to
+   compose them later or change grouped coordination yet.
+3. Preserve existing version checks for this refactor. Retain a small adapter
+   invoking current composition semantics at the existing lifecycle point.
+   Preserve source-version callbacks at acquisition, including header consumption.
+4. Run existing acquisition, dependency scheduling, item placement, source-version,
+   ordinary/layered equivalence, resource-lifetime, and source-context tests. Add
+   only targeted regressions for any newly split handoff not already covered.
+
+Validation: the focused suites above and the everyday check. This refactor must
+not change source IDs or turn acquisition order into precedence order.
+
+## Stage 10 — Separate Underlay-Group Acquisition From Composition
+
+Scope: behavior-preserving group coordination in `load_request.go`, using the
+per-input handoff already prepared in Stage 9.
+
+1. Separate acquiring an underlay group from scheduling/composing its acquired
+   scopes. Preserve namespace reanchoring, shared inclusion/repetition state,
+   item-local helper scopes, source/context aggregation, and group accounting.
+2. Keep authored acquisition order: main and depth-first imports, each underlay
+   and input, overlay, then synthetic CLI records. Preserve underlay indexes and
+   current version-check timing in this refactor; policy-dependent layered
+   composition is activated only in Stage 15.
+3. Expose the common scope-composition handoff with inputs for accumulated lower
+   policy and compiled application constraints. Initially adapt it to the current
+   semantics. Do not add another loader or repeat acquisition/version callbacks.
+4. Test the split with existing grouped/dependency/version/destination coverage,
+   including empty inputs/groups and repeated sources within versus across tiers.
+   Confirm unchanged values, source ordering, contexts, and failure evidence.
+
+Validation: grouped/layered, import scheduling, version, destination, and
+resource-context tests plus the everyday check.
+
+## Stage 11 — Compose Strict Collections With Resolved Identity
+
+Scope: the policy-aware strict composition core. Nested pending-item scope
+integration is deferred to Stage 12, not absorbed into this prompt.
+
+1. Extend `composeDocumentsWithPolicy` and `validateRegisteredSequences` to use
+   normalized field sets and scalar tuples. Resolve current declarations,
+   applicable application constraints, and eligible lower identity before peer
+   composition, using the Stage 4 agreement and Stage 5 lifetime helpers.
+2. Use the existing dependency-ready order for contributions and authored list
+   order. Validate every identified sequence, including singleton contributions;
+   strictly append distinct peers and reject duplicates under all modes.
+3. For explicit replacement, do not use obsolete lower YAML identity to authorize
+   otherwise unkeyed peer composition. Still apply independently registered
+   application constraints and genuinely applicable declarations.
+4. Attach Stages 7–8 evidence and Stage 5 scoped policy to composed results. Test
+   completed-document contributions with lower-established identity, declaration
+   conflicts, compound/scalar IDs, singleton validation, and replacement resets.
+
+Validation: composition and sequence identity tests plus the everyday check.
+Keep unregistered non-mapping conflicts unchanged. Public load activation remains
+in the later integration stages.
+
+## Stage 12 — Integrate Policy-Aware Item And Helper Scopes
+
+Scope: adapt pending-item/helper composition to Stage 11's core using the acquired
+scope boundaries from Stages 9–10.
+
+1. Gather declarations before composing each relevant scope, complete item-local
+   helper imports before item placement, and consume applicable non-main version
+   headers before checking record payload identity. Keep relative policy with its
+   owner until concrete placement is known.
+2. Apply eligible lower identity at each scope, but prevent a replacement outer
+   record from borrowing nested identity owned by the discarded lower record.
+   Preserve independently applicable policy and application constraints.
+3. Retained structural directives cannot serve as identity fields. A helper may
+   supply payload identity components. Do not require a helper-built compound
+   parent reference to be complete at this point; validate the completed surviving
+   reference during inheritance.
+4. Project item evidence and owned policies through Stages 7–8 machinery. Test
+   nested items, helper-supplied identity, header consumption, declaration scope,
+   discarded owner identity, and exact diagnostic paths with no value disclosure.
+
+Validation: item/destination/import/composition/source-version tests and the
+everyday check. Keep current public behavior until load-path activation.
+
+## Stage 13 — Activate Ordinary Source Collection Declarations
+
+Scope: ordinary `Load` base-source assembly, excluding activation of the new
+selected-overlay policy surface until Stage 14.
+
+1. Wire common directive extraction into every ordinary source form: file,
+   native filesystem, bytes, reader, mapping, imported mapping, and sequence item.
+   Declarations work without `Imports`; disabled imports retain their existing
+   rejection semantics. `NoSource` remains absence rather than an empty document.
+2. Use Stages 11–12 composition and Stage 5 scoped policy. Strip collection
+   controls from application decoding, but retain their locations/evidence for
+   their structural lifetime and diagnostics.
+3. Integrate policy/projections together with the values. Cover inline/imported
+   equivalence, placement-relative declarations, singleton and peer validation,
+   and declarations without accompanying values. Maintain ordinary loading's
+   existing non-collection overlay behavior while preparing its policy handoff.
+4. Update the implemented declaration/composition contract sections. Collection
+   precedence, inheritance, and later CLI behavior remain incomplete until their
+   integration stages; do not claim them implemented here.
+
+Validation: ordinary-load, import, composition, and metadata tests plus
+the everyday check. Include one public ordinary-load declaration example test.
+
+## Stage 14 — Activate Selected Overlay Collection Policies
+
+Scope: ordinary loading's selected overlay and shared precedence integration,
+using the source/declaration handoff already active in Stage 13.
+
+1. Permit `configkit.sequences` in selected overlays while continuing to reject
+   overlay imports. Preserve source-version header/control handling and ordinary
+   overlay namespace selection. Consume controls before application decoding.
+2. Use Stage 6's sequence operations and Stages 4–5 policy helpers in the assembly
+   accumulator. Mappings still recurse and unrelated sequences still replace.
+   Wire value, policy retirement/transfer, and Stages 7–8 inspection projections
+   together, not as a later best-effort repair pass.
+3. Cover append/upsert/replace, empty/null/kind cases, key-only/mode-only changes,
+   retained explicit modes, overlay identity agreement, and independently owned
+   policies. Check values, history, origins, and discarded descendants together.
+4. Update the selected-overlay and collection-operation contract sections. Make
+   the shared precedence entry reusable by layered loading without adding another
+   operation implementation or turning overlay imports on.
+
+Validation: ordinary-load, overlay, composition, policy, and metadata tests plus
+the everyday check.
+
+## Stage 15 — Integrate Lower Policy Across Layered Assembly
+
+Scope: source-backed underlay coordination and the layered selected-overlay path,
+using the prepared acquisition, ownership, strict-scope, and precedence machinery.
+Do not absorb unfinished work from Stages 5–14 into this stage.
+
+1. Acquire main first, but assemble underlays from lowest to highest, strictly
+   composing each group with applicable accumulated lower policy before applying
+   its one precedence operation. Then compose/apply main and the selected overlay.
+   Do not sequentially upsert peer sources as if they were different tiers.
+2. Pass lower identity into higher strict scopes before peer composition. Cover an
+   otherwise empty underlay establishing identity, inherited explicit mode, and
+   replacement that deliberately discards the old YAML key. Invalid lower tiers
+   still fail even if a higher replacement would hide their values.
+3. Keep main-only version ownership and every enabled per-source version rule.
+   In default source-version mode, run the main checker once on the composed
+   main tier before overlay/inheritance/defaults, never on underlay-augmented
+   values. Its former main-before-underlay composition timing need not persist:
+   main composition now depends on lower policy. Enabled-mode source checks keep
+   acquisition order and exact source-local header handling. No extra callbacks
+   or second acquisition pass are permitted.
+4. Keep histories in precedence order, sources in acquisition order, underlay
+   indexes stable across empty groups, and shared budgets intact. Update the
+   layered contract's acquisition-versus-composition description accordingly.
+   Remove the acquisition/composition adapters superseded by this integration;
+   the old public-registration adapter remains until Stage 22.
+5. Cover multiple inputs/groups, mounted main inputs, item scopes, empty main,
+   explicit empty groups, layered overlays, version protection and diagnostics,
+   and equivalence with ordinary loading where the inputs express the same model.
+
+Validation: layered/grouped, source-version, dependency scheduling, destination,
+and metadata tests plus the everyday check.
+
+## Stage 16 — Verify Replacement And Nested Policy Ownership End To End
+
+Scope: the consequential lifetime cases across the now-active source assembly.
+
+1. Exercise an item-local nested collection through retaining the outer record,
+   upserting the same outer ID, and replacing the entire outer sequence. Retained
+   items keep their local key/mode; replacements do not inherit discarded policy
+   merely because identity, source label, or final index is equal.
+2. Exercise ancestor replacement and later reintroduction. Removed collection
+   identity does not reappear; independently declared append mode can survive
+   and require renewed identity. Cover renewed YAML key, application constraint,
+   explicit replace, and the intended failure with no available identity.
+3. Cover key-only validation of retained records, pending declarations, a no-value
+   replace declaration that cannot authorize key changes, and independent policy
+   whose owning scope survives a record replacement.
+4. Correct any integration defects in the Stage 5 ownership transformations;
+   consolidate tests with their mechanism owners. Do not introduce special-case
+   ownership by numeric path or a historical identity cache to make cases pass.
+5. Verify frozen metadata, origin/resource context, sensitivity of discarded
+   records, and mutable independence across repeated loads for these scenarios.
+
+Validation: focused policy/sequence/layered/ownership tests,
+`go test -race ./internal/engine`, and the everyday check.
+
+## Stage 17 — Prepare Inheritance Policy And Template Transport
+
+Scope: behavior-preserving inheritance state transport, before activating
+collection-aware inheritance operations in Stage 18.
+
+1. Extend inheritance templates/expansion sidecars to carry collection identity
+   and record-local policy alongside nodes and origins. Keep existing scheduler
+   keys, cancellation, version protection, missing-parent/cycle diagnostics, and
+   destination-local evaluation of nested templates.
+2. Provide owned-state copy/transfer helpers from Stage 5 and path projections from
+   Stage 8. Preserve original nested instructions for each destination's later
+   evaluation rather than retaining only a parent's expanded output. Keep
+   application registration matching separate from copied YAML policy.
+3. Test template reuse at multiple destinations, independent scoped state, and
+   retained instruction locations/resource context. Preserve current inheritance
+   values/history while these new helpers remain preparatory; do not activate a
+   value change without its corresponding policy/evidence handling.
+
+Validation: existing inheritance, ownership, history, and resource-context tests,
+focused transport tests, and the everyday check.
+
+## Stage 18 — Apply Collection Policy During Mapping-Sibling Inheritance
+
+Scope: activate shared collection operations using the prepared template/state
+handoff; existing mapping-sibling relationship selection remains unchanged.
+
+1. Use the shared sequence combiner at the inheriting destination: parent lower,
+   child higher, destination explicit mode or identity-based default. Copy YAML
+   identity where applicable, but do not automatically copy a parent collection's
+   explicit mode to a different destination. Match application constraints at the
+   destination, including newly created matching paths.
+2. Transfer record-owned state only with surviving records using Stage 17's
+   helpers. Integrate inheritance evidence with Stage 8's projections in the same
+   activation, preserving origins, historical sensitivity, resource context,
+   cancellation, and existing expansion/history accounting.
+3. Cover nested append/upsert/replace, explicit empty/null/kind cases, copied key
+   with inferred upsert, destination explicit mode, incompatible keys, application
+   constraints, cancellation, and policy at an inherited new namespace.
+4. Update current mapping-sibling inheritance documentation for collection
+   operations. Identified-record parent lookup remains incomplete until Stage 21.
+
+Validation: inheritance, history, provenance, sensitivity, and resource-context
+tests plus the everyday check. No identified-record parent lookup is active yet.
+
+## Stage 19 — Build Identified-Record Inheritance Graphs
+
+Scope: relationship extraction/validation and shared scheduling for sequence
+scopes; test these helpers before activating payload expansion.
+
+1. Extend sibling-scope discovery to identified sequences. Build the tuple-to-item
+   lookup once using effective identity. Every record supplies its own complete
+   identity before inheritance; neither parent lookup nor typed defaults supplies
+   a missing child field.
+2. Parse a surviving record's `extends` as null cancellation or a mapping with
+   exactly the normalized identity fields and supported scalar components.
+   Mapping-sibling references keep their existing string form. Do not accept a
+   scalar shorthand, multiple parents, positional or cross-collection lookup.
+3. Parent-reference mappings have already used ordinary strict mapping
+   composition: disjoint helper fields can complete one reference, while equal
+   overlapping fields conflict. Validate completeness/extra fields only now on
+   the surviving reference. Discarded references remain historical evidence.
+4. Build one-parent dependency edges and reuse `scheduleDependencies` with stable
+   semantic-pointer keys. Missing parents, self-reference, and longer cycles
+   use existing inheritance diagnostics without identity values. Keep authored
+   output positions independent from graph schedule.
+5. Cover compound and null references, numeric equality, forward edges, chains,
+   unrelated equal identities in different lists, malformed refs, missing parents,
+   cycles, helper-built references, and overwritten/cancelled relationships.
+
+Validation: focused inheritance graph/reference tests and the everyday check.
+
+## Stage 20 — Prepare Reusable Inheritance Scope Execution
+
+Scope: remove mapping-container assumptions from the private template executor
+without activating identified-record inheritance in public loads.
+
+1. Refactor template execution to separate relationship scheduling/payload
+   expansion from reading and rebuilding its scope container. Preserve current
+   mapping-sibling output, locations, cancellation, evidence, and budgets.
+2. Provide private mapping- and identified-sequence-scope adapters using Stage 19's
+   relationship graphs. The sequence adapter preserves output positions and
+   exposes child identity fields to the payload-expansion step. Keep this private;
+   do not add a public generic inheritance or merge callback API.
+3. Reuse Stage 17's template/policy transport and Stage 18's collection-aware
+   payload operations. Existing origin/history/retained-node handling must remain
+   available to both adapters rather than duplicating an entire executor.
+4. Run existing mapping-inheritance regressions and privately exercise a small
+   sequence graph through the adapter. Check output order, independent templates,
+   and operation/evidence paths before public activation in Stage 21.
+
+Validation: inheritance execution, scheduling, history, and budget tests plus
+the everyday check. This refactor does not introduce new parent-reference forms
+or change public loading semantics yet.
+
+## Stage 21 — Resolve Record Inheritance With Ownership And Evidence
+
+Scope: integrate Stage 19's graphs and Stage 20's executor into the structural
+expansion phase, with values and inspection working together.
+
+1. Expand parent payload into the child while preserving the child's own identity
+   fields and their origin/presence. Exclude parent identity fields from payload
+   inheritance. Recursively combine mappings and use destination collection policy
+   for nested sequences; retain whole-record upsert at the earlier boundary.
+2. Consume the child's directive before decoding. Keep parents as ordinary records
+   subject to the same schema/semantic validation; create no abstract templates or
+   acquired source records.
+3. Resolve only records surviving tier assembly. Cover parent upsert/removal,
+   replacement of a child without `extends`, self-reference not exposing a lower
+   predecessor, and nested destination-local relationships. Enclosing scopes
+   finish before nested scopes; assembled list order stays unchanged.
+4. Carry policy, origin/history, sensitivity, resource context, and existing
+   expansion/retained-history accounting through copies. Do not widen budgets or
+   require a new inspection-path API.
+5. Add public-load regressions for complete inheritance chains and nested
+   collections, including secret canaries on success and failure. Update the
+   implemented inheritance contract and consumer guidance for this scope.
+
+Validation: all inheritance and cross-feature inspection tests,
+`go test -race ./internal/engine`, and the everyday check.
+
+## Stage 22 — Expose The Root-Owned Registration API
+
+Scope: the settled facade surface and migration of in-repository callers.
+
+1. Implement the exact public names/signatures above in a focused root file, and
+   add `Definition.SequenceIdentities`. Convert root records/selectors explicitly
+   into the engine model. Snapshot fields, selectors, and paths on construction
+   and loading, preserving immutable definition reuse.
+2. Remove `SequenceComposition`, `Definition.SequenceCompositions`, and the private
+   old-shape adapters. Migrate all maintained code/tests/examples; do not edit
+   historical release notes solely to erase the old name or add deprecated aliases.
+3. Cover mixed exact/dynamic constructors, compound scalar fields, literal `*`
+   versus wildcard segments, root/unused selectors, invalid index paths and empty
+   field sets, overlap failure before source access, and independent loads.
+4. Prove dynamic constraints apply before strict composition and at inherited
+   destinations, without consumer namespace enumeration. Keep final-result
+   enforcement described as pending until Stage 27 actually adds it.
+5. Verify the public boundary and declaration documentation. Do not expose engine
+   enums, policy stores, parser nodes, or internal methods.
+
+Validation: root sequence/facade, registration, public/dependency-boundary, and
+consumer example tests plus the everyday check.
+
+## Stage 23 — Separate Reusable CLI Override Preparation
+
+Scope: mechanical preparation refactor and a privately tested post-structural
+validation entry, without changing current CLI processing order.
+
+1. Refactor override preparation so labels, paths, snapshots, type support,
+   disjointness, version protection, and existing budgets remain reusable outside
+   the source assembly accumulator. Allocate synthetic sources once in the same
+   order; do not reopen or reinterpret CLI values as YAML.
+2. Prepare post-structural validation that rejects `extends` as well as
+   `configkit` in supplied paths and nested values, using their established
+   structural roles. Test this entry privately; retain the existing entry's
+   behavior until the lifecycle switches in Stage 26. Do not reject previously
+   valid early-CLI instructions before that switch or add unrelated restrictions.
+3. Test detached prepared batches, synthetic source ordering, and the new
+   directive validation. Reuse existing label/path/cycle/type-support/budget tests
+   rather than adding another inventory of the same preparation cases.
+
+Validation: override preparation/source tests and the everyday check. Keep
+ordinary/layered behavior unchanged; typed assignment is the next stage.
+
+## Stage 24 — Implement Schema-Aware Post-Decode CLI Assignment
+
+Scope: a private typed assignment helper using Stage 23's prepared overrides,
+without activating the later lifecycle order yet.
+
+1. Add `applyDecodedOverrides`, using existing schema/decode mechanisms to bind
+   only supplied override subtrees against a detached copy of the decoded state.
+   Mappings recursively overlay, sequences replace, and explicit zero/empty/null
+   retain normal schema rules. Current typed values supply retained children,
+   not new defaults for replaced list items. Do not call `Definition.Defaults`
+   again or re-decode the whole projected configuration as newly authored input.
+2. Return the updated typed value, canonical post-CLI projection, presence/origin
+   changes, and existing raw CLI history contributions. Preserve a pre-CLI typed
+   projection for decoding reconciliation/history. No resolver runs on a failed
+   override; no partially applied result is returned.
+3. Test unknown/type-invalid fields, mapping retention, sequence replacement,
+   nested maps/lists, named/text/duration/secret schema fields, explicit null/false/
+   zero, discarded children, structural controls, and mutable ownership.
+
+Validation: override, decode, schema, projection, and ownership tests plus the
+everyday check. Only the helper is added here; the shared lifecycle does not
+switch until its inspection/context support is prepared in Stage 25.
+
+## Stage 25 — Prepare Post-Decode CLI Inspection And Effective Context
+
+Scope: prepare history, resolution-input, and secret-lookup context handling;
+retain the existing public CLI lifecycle until Stage 26.
+
+1. Extend existing replay/reconciliation to represent the target order:
+   source tiers, inheritance, typed decoding/defaults, CLI, application resolution.
+   Keep the structural input, pre-CLI typed snapshot, raw CLI contributions, and
+   post-CLI projection distinct where needed. Use the existing history machinery,
+   not a second replay framework, and retain legacy input compatibility until the
+   lifecycle switch. Do not mislabel CLI changes as defaults or inheritance.
+2. Prepare resolution-entry metadata from Stage 24's result so explicit presence,
+   origins, source records, and resource contexts describe the overridden state.
+3. Prepare the effective reference/location handoff for secret lookup and final
+   validation using the post-override/post-resolution values, not stale structural
+   input. Preserve existing no-credential-byte and ownership behavior.
+4. Test the target pipeline with privately assembled snapshots: equal-value CLI
+   replacement, replaced descendants, decoding reconciliation before CLI, later
+   recorded resolver transformations, explicit-presence defaults, and an overridden
+   secret reference. Verify current public-load regressions still pass unchanged.
+
+Validation: history/metadata/resolution/secret tests and the everyday check.
+Public activation must not require a later stage to repair these mechanisms.
+
+## Stage 26 — Activate CLI After Decoding In The Shared Lifecycle
+
+Scope: activate ADR 0018 through both loading paths using Stages 23–25's complete
+assignment, inspection, and effective-value handoffs.
+
+1. Remove CLI assignment from ordinary/layered structural assembly. Invoke
+   Stage 24 after inheritance and successful typed-default creation/decoding,
+   immediately before application resolution. Activate Stage 23's post-structural
+   validation. Keep the single shared lifecycle, not separate CLI implementations.
+2. Wire Stage 25's history and context handoffs in the same change. Resolution
+   receives explicit-presence/origin metadata and effective CLI values;
+   `Resolution.Default` preserves them while recorded transformations can change
+   them normally. Secret lookup uses effective references and locations. Preserve
+   exactly-once default-factory behavior and synthetic source ordering.
+3. Cover `NoSource`, ordinary/layered equivalence, CLI clearing/replacing an
+   identified list despite append/upsert mode, derived values seeing CLI input,
+   equal-value provenance, and resolver changes after CLI. Add explicit regressions
+   that CLI cannot fix earlier decode/inheritance errors or introduce/cancel a
+   relationship. Update obsolete early-CLI tests deliberately.
+4. Update the lifecycle, override, history, and secret-reference contracts and
+   affected examples to describe the now-implemented order.
+5. Remove the superseded early-CLI application paths, old preparation entry, and
+   temporary replay compatibility paths. Keep their still-relevant behavioral
+   tests under the new lifecycle rather than postponing code cleanup to the
+   documentation stage.
+
+Validation: override/lifecycle/resolution/inheritance/history/secret tests,
+`go test -race ./...`, and the everyday check.
+
+## Stage 27 — Enforce Final Application Identity Constraints
+
+Scope: the final-result guarantee, not another combination or inheritance phase.
+
+1. Add shared final identity validation after effective secret lookup and before
+   application semantic validation/publication in the common lifecycle. Use the
+   canonical effective typed projection, which retains secret references rather
+   than credential bytes. Preserve the existing validated publication snapshot
+   and detachment behavior.
+2. Traverse the actual final tree with the compiled selector trie, validating each
+   matching sequence independently with Stage 1 primitives. This must discover
+   paths introduced by typed defaults, CLI, and callbacks, not only assembly paths.
+   Support root/namespaced rules and unrelated unselected sequences.
+3. Validate actual sequences only. Missing paths, null, and other shapes remain
+   the schema/application validator's responsibility. Identity extraction here
+   examines effective typed payload, not whether a field was authored. Do not
+   recheck YAML-only policies or every intermediate callback state.
+4. Cover duplicate/missing/non-scalar identities produced at each late phase,
+   valid null/scalar identities, numeric equality after typed conversion, dynamic
+   namespaces created by each late phase, harmless unmatched rules, concrete-path
+   diagnostics with available origins, and secret-value non-disclosure.
+5. Prove both `Load` and `LoadLayered` use this guarantee and return no usable
+   result on failure. Update registration comments and canonical contract with
+   this final boundary, excluding later caller mutations from the guarantee.
+
+Validation: final identity, selector, facade, defaults/resolution/override, and
+secret tests plus the everyday check.
+
+## Stage 28 — Exercise Representative Consumer Workflows
+
+Scope: a small number of meaningful integrations, not a Cartesian test matrix.
+
+1. Extend `examples/consumers` with an executable source-backed catalog: underlay
+   records, main fragments, selected overlay append/upsert/replace, and explicit
+   record inheritance. Demonstrate complete replacement of a matched record,
+   stable order, and final inspection without application merge code.
+2. Add a namespaced typed-map example using one wildcard registration plus an
+   exact root registration, with independent uniqueness. Show YAML-declared
+   identity without registration separately so the final guarantee is not implied.
+3. Include one integrated CLI/resolution case and one representative provenance/
+   resource-context/secret-history case where those interactions add protection
+   beyond mechanism tests. Keep application domain validation recognizable.
+4. Audit coverage against the roadmap's work inventory, especially helper-built
+   parent references, version-header consumption, pending declarations, lower
+   identity before higher peers, replacement ownership, and dynamic destinations
+   introduced through inheritance and late typed changes. Add only missing,
+   consequential regressions at their narrowest stable owners.
+5. Consolidate obsolete/brittle tests. Do not restore old replacement semantics
+   merely to keep fixtures unchanged, and do not preserve duplicates of the same
+   semantic table across every layer.
+
+Validation: `go test ./examples/consumers`, focused root integration tests,
+`go test -race ./...`, and the everyday check.
+
+## Stage 29 — Reconcile Documentation
+
+Scope: finish the current-contract narrative. Code cleanup belongs with the
+integrations that supersede each adapter, not in this documentation prompt.
+
+1. Reconcile `docs/config.md` for declarations, identity equality, selectors,
+   operations/persistence, ownership, inheritance, final validation, and CLI order.
+   Keep its operation and presence tables authoritative; Go comments own exported
+   symbols and consumer guidance links rather than copying full contracts.
+2. Update `docs/consumers/README.md`, relevant root examples, `README.md`,
+   `docs/internal/overview.md`, `docs/internal/imports.md`, and
+   `docs/development.md` where the implemented structure or reading/test map has
+   changed. Remove stale claims about string-only registration, universal sequence
+   replacement, early CLI, and overlay prohibition of all directives.
+3. Remove implemented portions from the deferred identified-record discussion in
+   `future.md`, retaining genuinely deferred recursive body merging and other
+   still-relevant ideas. Preserve accepted ADR decision content; update only status
+   or implementation references where appropriate. Architecture already adopts
+   the decisions: remove pending-implementation wording when true, without adding
+   new policy. Leave release version selection/publication to the release task.
+4. Search maintained prose/examples for obsolete registration names and lifecycle
+   claims; retain intentional historical references. Follow canonical ownership
+   rather than copying the complete contract into each guide.
+
+Validation: local documentation-link checks and whitespace review; run relevant
+example tests if executable examples changed. Reuse Stage 28's successful
+code/race validation when its covered content is unchanged.
+
+## Stage 30 — Audit Scaling And Verify The Completed Target
+
+Scope: a bounded implementation review and final validation, not another broad
+refactor or expansion of the feature/performance scope.
+
+1. Inspect the completed code for duplicate combination/identity algorithms,
+   per-record scans of whole collections, repeated declaration gathering, numeric
+   magnitude expansion, stale ownership/path caches, and unnecessary copies.
+   Prefer shared indexed mechanisms without pooling/global caches. Reuse existing
+   composition/underlay/inheritance benchmarks and add a focused mixed-upsert or
+   wildcard-collection benchmark only where existing fixtures cannot exercise the
+   new scaling risk. Record commands and qualitative findings here; timing or
+   allocation thresholds and a quiet-host benchmark campaign are not gates.
+2. Confirm transitional adapters/dead early-CLI paths have been removed. Search
+   current code and maintained examples for `SequenceCompositions` and
+   `SequenceComposition`; only intentional historical references may remain.
+   Confirm no new dependencies, public internal-type exposure, background work,
+   release machinery, or additional limits were introduced. Review the full diff
+   against the roadmap, not merely passing tests. Record bounded review findings
+   here; significant corrective
+   work requires its own focused unit rather than being hidden inside this stage.
+   Do not declare completion while a target behavior remains incorrect or missing.
+
+Validation: `./scripts/check.sh`, `go test -race ./...`, and the development
+guide's local documentation/public-boundary checks, reusing successful checks
+from Stages 28–29 if their covered content is unchanged. Record completed scope,
+validation, benchmark commands/qualitative findings, and any real remaining
+issues accurately. No new timing thresholds or release gates are introduced.
+
+Acceptance: all roadmap operations and lifecycle boundaries work through ordinary
+and layered loads; values and inspection agree; application constraints hold on
+successful final results; deferred behavior and YAML-only guarantees have not
+expanded. All thirty stages are complete and their nonempty units committed.
+
+## Open Questions
+
+None. Feature choices are settled in the roadmap and accepted ADRs; this plan
+selects the remaining API, representation, validation, and execution decisions.

docs/roadmap/sequences.md

@@ -0,0 +1,999 @@
+# Sequence Identity And Combination Roadmap
+
+Status: settled feature roadmap; implementation is specified in the
+[staged implementation plan](implementation.md). This document defines the
+targeted end state, not an implemented contract. The implementation plan owns
+execution order, exact Go names and signatures, bookkeeping, and validation.
+
+## Purpose And Scope
+
+Allow configuration authors to assemble collections from independently authored
+records, declare identity in configuration, and progressively extend a baseline
+collection through higher layers. Authors can replace the entire collection,
+append distinct records, or replace complete records by identity while retaining
+unmatched records. Sibling inheritance uses the same sequence operations and
+identity-based defaults. Within an identified sequence, a record can explicitly
+inherit settings from another record in that sequence while keeping its own
+identity.
+
+The motivating use case is a prompt catalog whose source-backed baseline supplies
+records and whose application or installation layers add or supersede records.
+Applications can optionally impose identity constraints that configuration must
+respect and that hold in the final loaded configuration. These constraints can
+cover exact destinations or collections beneath dynamically named mapping keys,
+without requiring the application to discover those names before loading.
+
+The feature covers ordinary and layered loading, imported items, inline
+sequences, selected overlays, and sibling inheritance. Typed defaults, CLI
+assignments, and application resolution retain their own combination semantics;
+CLI application moves later in the loading lifecycle.
+
+This scope is narrower than the deferred
+[identified-record sequences proposal](future.md#1-identified-record-sequences).
+Automatic recursive merging of matching records remains outside this scope;
+explicit inheritance between distinct records is included. The initial
+declaration surface is `configkit.sequences`; `key` and `mode` shorthand on
+imports is deferred. Application lists remain
+ordinary YAML sequences without control wrappers. Unkeyed sequence concatenation
+is outside this feature.
+
+## Current Baseline And Required Changes
+
+The [import contract](../config.md#explicit-imports) owns existing selectors,
+destinations, placement, and ordering. `placement: sequence-item` already places
+each selected source mapping as one item at its effective destination. Item-local
+helper imports can contribute fields before placement.
+
+The [composition contract](../config.md#strict-document-composition),
+[layered-loading contract](../config.md#layered-loading), and
+[inheritance contract](../config.md#sibling-inheritance) describe the implemented
+baseline:
+
+| Area | Current behavior | Target behavior |
+| --- | --- | --- |
+| Identity declaration | `Definition.SequenceCompositions` registers one literal string field at each registered mapping path. | YAML collection declarations and optional authoritative application registrations support field sets and scalar identity components. |
+| Application registration destinations | Registrations select exact literal mapping paths. | Exact and structural selectors can impose identity beneath arbitrary mapping keys, independently at each matched destination. |
+| Strict composition | Registered sequences append distinct string identities and reject duplicates; repeated unregistered sequences conflict. | Collections with effective identity append distinct records and reject duplicates. Applicable lower-tier identity is available before composing higher-tier peers. Unidentified sequences retain strict-conflict behavior. |
+| Source precedence and selected overlays | A present higher sequence replaces the lower sequence wholesale. | Collection policy selects `replace`, `append`, or `upsert`; overlays can declare policy for inline sequences. |
+| Sibling inheritance | An inheriting mapping recursively overlays its resolved parent; supplied sequences replace wholesale. | Mappings retain recursive combination, while sequences use destination policy and the same three operations. |
+| Sequence-member inheritance | A sequence element has no sibling scope for its own inheritance relationship. | An identified sequence provides a sibling scope: a record can inherit from one record in the same sequence, selected by complete identity. |
+| CLI overrides | Apply before inheritance and typed defaults/decoding. | Apply after inheritance and typed defaults/decoding, before application resolution. |
+| Identity enforcement | Registration governs strict sequence composition. | Structural identity checks cover identified combinations; application registrations additionally constrain final effective sequences. |
+
+Source-backed underlays participate in source assembly. `Definition.Defaults`
+provides typed defaults later; extending source combination does not turn typed
+default lists into source underlays.
+
+## Collection Declarations
+
+Use a source-root `configkit.sequences` block containing collection declarations.
+Each entry identifies a `destination` and may declare `key`, `mode`, or both.
+Identity and mode can be supplied independently. This form serves imported items,
+inline records, ordinary mapping imports, and inline sequences in selected
+overlays.
+
+```yaml
+configkit:
+  sequences:
+    - destination: [prompts]
+      key: [namespace, id]
+      mode: append
+  imports:
+    - dir: prompts
+      pattern: '*.yaml'
+      destination: [prompts]
+      placement: sequence-item
+```
+
+Each selected file contributes a record such as:
+
+```yaml
+namespace: support
+id: welcome
+text: Welcome
+```
+
+Here append retains a lower-layer collection and adds distinct higher-layer
+records. Identity and combination controls belong to the collection declaration,
+so imported and inline contributions share one policy surface.
+
+### Destinations And Scope
+
+`destination` uses the existing sequence of literal mapping keys. Resolve it
+relative to the declaring source's placement: a fragment mounted under `catalog`
+can declare `[prompts]` locally. A sequence-item source can declare a nested
+collection within its own item without referring to an eventual numeric index.
+Preserve source and item-local scope boundaries.
+
+A collection's scope is its resolved destination within its composition scope.
+Resolve relative destinations before reconciling declarations. Different imported
+items retain separate scopes for their nested collections. A declaration governs
+all contributions to its collection in the tier, including imported and inline
+records; it does not govern only records in the declaring file.
+
+Gather declarations before composing the relevant scope so their effect does not
+depend on which contribution is assembled first. Retain applicable controls
+through source precedence and inheritance, then keep them out of application
+decoding.
+
+Sequence declarations are independent of filesystem import enablement. Selected
+overlays may contain `configkit.sequences`, extending the current blanket
+prohibition on overlay `configkit` declarations. This does not enable imports
+in overlays.
+
+### Overlay Authoring
+
+For an existing sequence `[foo, bar]`, an overlay can replace `foo` in full:
+
+```yaml
+configkit:
+  sequences:
+    - destination: [prompts]
+      key: id
+      mode: upsert
+prompts:
+  - id: foo
+    text: Revised definition
+```
+
+The result is `[foo*, bar]`. The explicit upsert mode can be omitted when no
+different explicit mode applies from a lower tier: effective identity then
+selects upsert. To add a distinct record instead:
+
+```yaml
+configkit:
+  sequences:
+    - destination: [prompts]
+      key: id
+      mode: append
+prompts:
+  - id: baz
+    text: Another prompt
+```
+
+The result is `[foo, bar, baz]`. To replace the whole sequence:
+
+```yaml
+configkit:
+  sequences:
+    - destination: [prompts]
+      mode: replace
+prompts:
+  - id: bar
+    text: Bar definition
+  - id: baz
+    text: Baz definition
+```
+
+The result is `[bar, baz]`. Explicitly selecting the default operation is valid.
+With no applicable explicit mode, omitting this declaration selects replace for
+an unidentified collection and upsert for an identified one. A retained explicit
+mode continues to govern.
+
+## Identity Definitions And Constraints
+
+### Field Sets
+
+`key` accepts a string or a nonempty list of strings naming direct, literal
+fields in each record. A name such as `metadata.id` denotes that literal field,
+not a nested lookup. Treat field names as a set: remove repeated names and
+ignore authored order when comparing definitions.
+
+Thus `id`, `[id]`, and `[id, id]` are equivalent; `[namespace, id]`,
+`[id, namespace]`, and `[namespace, id, id]` are also equivalent. An empty list
+is invalid because it supplies no identity fields. These rules apply to both
+configuration declarations and application registrations.
+
+Use a consistent field order internally to construct identity tuples. Preserve
+component types and tuple boundaries so unstructured concatenation cannot cause
+otherwise distinct identities to collide.
+
+### Scalar Components And Equality
+
+Each identified record must be a mapping with every designated identity field
+present. Components may be strings, numbers, booleans, or explicit null. Missing
+fields and mapping or sequence components are invalid. Present null differs
+from absence; two single-field null identities are equal. Compound identities
+are equal when all corresponding components are equal.
+
+Strings compare exactly, including empty strings, without trimming, case folding,
+or normalization. Scalar categories remain distinct: `"1"`, `1`, `true`, and
+null are different identities.
+
+Numbers compare by mathematical value across integer, fractional, and exponent
+spellings: `1`, `1.0`, and `1e0` are equal, as are `0` and `-0`. Use exact numeric
+normalization for identity comparison without rounding through a floating-point
+conversion. This does not rewrite authored values or change existing typed
+decoding and projection semantics. The parser's rejection of non-finite numeric
+forms remains unchanged.
+
+### Application-Owned Constraints
+
+Retain an optional application interface for registering collections and imposing
+their identity definitions. Configuration may define identity where no application
+constraint exists. When a declaration omits `key`, an applicable registration
+supplies identity. An explicit equivalent definition is accepted, and a different
+definition fails under every mode, including replace.
+
+For example, if the application requires `id`, `key: [id, name]` is invalid:
+that compound identity would allow repeated `id` values with different names.
+A registration is an authoritative constraint, not an omission-filling default.
+Copying a collection through inheritance does not create another application
+registration; registrations constrain their registered destinations.
+
+Identity constraints validate actual sequences. They do not independently require
+a path to exist, be non-null, or have sequence shape. The application schema and
+validation own those requirements. Final-result enforcement for application
+registrations is defined under [Validation And Diagnostics](#validation-and-diagnostics).
+
+### Application Registration API
+
+Use one public sequence-identity registration type, collected in one field on
+`Definition`. Each registration contains an explicit destination selector and
+its identity fields. Exact and dynamic registrations share matching, overlap
+validation, identity comparison, and lifecycle behavior.
+
+Provide two constructors returning that same registration type:
+
+- An exact-path convenience constructor accepting a literal `Path` and identity
+  field names. It converts the path into a selector containing literal segments.
+- An explicit-selector constructor accepting a structural selector and identity
+  field names, for destinations that include arbitrary mapping keys.
+
+Both constructors accept identity fields as variadic strings, so one-field and
+compound identities use the same construction form. Apply the settled field-set
+rules, including deduplication. Invalid registrations, including an empty identity
+field set, fail during application-definition validation.
+
+Existing `Path` values retain their literal meaning throughout ConfigKit. The
+exact-path constructor supplies convenience at the registration boundary; dynamic
+matching uses the explicit selector type. Both forms can appear together in the
+same definition field.
+
+Replace the current registration shape with this model. Preserving the existing
+`SequenceComposition` fields and `Definition.SequenceCompositions` spelling is not
+a requirement, and no compatibility facade is needed solely to retain that syntax.
+Choose names centered on sequence identity, reflecting its role in composition,
+inheritance, and final validation. Exact Go type, field, and constructor names
+and signatures belong in the implementation plan.
+
+### Dynamic Application Registration Destinations
+
+Allow applications to register identity for collections whose containing mapping
+keys become known only during loading. For example:
+
+```yaml
+collections:
+  weatherreporter:
+    prompts:
+      - id: foo.bar
+        text: Report the weather.
+  homewatch:
+    prompts:
+      - id: foo.bar
+        text: Report the household status.
+```
+
+One application registration can require `id` for every namespace's `prompts`
+list. The equal IDs above are valid because each list has its own uniqueness
+scope. An optional root `prompts` list can have a separate exact registration.
+
+Use an explicit application-side structural selector composed of literal mapping
+keys and an arbitrary-mapping-key segment. Conceptually, the namespace rule is:
+
+```text
+Literal("collections") -> AnyMappingKey -> Literal("prompts")
+```
+
+This notation explains the selector; it is not a Go signature or YAML syntax.
+Each segment consumes exactly one mapping level. Match the complete destination,
+not an arbitrary prefix or suffix. Literal keys remain exact, including empty
+strings and keys containing punctuation such as `*`, `/`, or `.`. An arbitrary-key
+segment is distinct from any literal spelling and does not match a sequence
+index. Existing `Path` values retain their literal meaning.
+
+This capability belongs to the reusable application definition and applies to
+ordinary and layered loads. YAML collection destinations and import destinations
+retain their existing literal syntax. Recursive search, regular expressions,
+filesystem glob rules, sequence-index selectors, and custom matching callbacks
+are outside this addition.
+
+At each concrete matched destination, apply the same authoritative identity
+constraint as an exact registration. Use the settled field-set and scalar-equality
+rules. Do not pool identities across destinations or require records to repeat
+the containing mapping key as an identity component. Applications retain ownership
+of namespace interpretation, qualified runtime IDs, and domain validation.
+
+Matching supplies effective identity before relevant strict composition and
+precedence operations. Without an applicable explicit mode, this identity selects
+upsert under the existing default rule. It also supplies identity for inheritance
+within the matched sequence. YAML declarations must agree with it under every
+mode, including replace. These are ordinary application constraints with a broader
+destination selector, not a separate final-validation facility.
+
+### Registration Overlap
+
+Equivalent application constraints coalesce at any shared destination. Equivalence
+uses the identity field-set rules: for example, `id` and `[id, id]` agree.
+Incompatible constraints fail; exact registrations do not override broader
+selectors, and registration order grants no precedence.
+
+Detect incompatible overlapping registrations while validating the application
+definition, before source acquisition. For these fixed-depth mapping selectors,
+overlap exists when the selectors have the same number of segments and every
+corresponding pair can match a common key. Unequal literal keys at any position
+make the selectors disjoint; an arbitrary-key segment can overlap a literal or
+another arbitrary-key segment. Compare segment structure rather than rendered
+path strings. No configuration pre-read is needed.
+
+Thus an `id` rule for all namespace prompt lists conflicts with a `[id, name]`
+rule for one named namespace, even if that namespace is absent from a particular
+load. This is an invalid application definition. A valid selector matching no
+collections is harmless: registration does not require a path to exist, create
+collections, or impose shape and nullability requirements.
+
+### Matching Across The Loading Lifecycle
+
+Match application selectors against effective configuration destinations after
+accounting for input placement. Physical filenames and source-local paths do not
+determine applicability. Make constraints available before composing contributions
+at those destinations, including namespaces first introduced by imports, mounted
+inputs, or higher tiers. Apply matching constraints to key-only YAML declarations
+as well, even before their sequence exists. Consumers need neither a preliminary
+load nor source inspection to enumerate namespaces.
+
+During inheritance, apply registrations at the inheriting destination, including
+newly created matching destinations. Copying a collection does not transfer its
+application registrations to a different path. The existing YAML-policy ownership
+and transfer rules remain separate from application selector matching.
+
+During final validation, find matching actual sequences in the final effective
+configuration. This includes namespaces introduced or replaced by typed defaults,
+CLI overrides, and application resolution. Reusing only destinations found during
+source assembly would miss these values. Use the same selector meaning and
+identity validator in both phases; later assignments do not trigger composition
+or inheritance again. YAML-only identities retain their existing enforcement
+boundary.
+
+## Strict Composition And Precedence Operations
+
+### Compose Peers Before Applying Precedence
+
+Resolve applicable identity and mode policy, strictly compose the tier's peer
+contributions, then apply one selected operation per collection against the
+accumulated lower value. The tier operation does not sequentially replace peer
+source contributions. Selected overlay contributions use the same combination
+semantics.
+
+Within a strict tier, equivalent identity definitions coalesce and different
+field sets fail. Agreeing explicit modes coalesce; conflicting explicit modes
+fail. Omission expresses no preference. A root declaration or a later declaration
+has no precedence over conflicting peers. Explicitly selecting a default mode
+is still a real instruction and can conflict with another explicit mode.
+
+Identified peer sequences append distinct records. Duplicate identities among
+peers or within a contributed list fail under all three modes, including replace
+and upsert. Equal or disjoint record payloads do not excuse a duplicate identity.
+Unidentified sequences retain their current strict-composition conflict behavior.
+
+For example, higher-tier fragments `[C]` and `[D]` with an agreed identity assemble
+as `[C, D]`. Against lower `[A, B]`, append or upsert produces `[A, B, C, D]`,
+while replace produces `[C, D]`. Peers `[B]` and `[B*]` conflict even under
+upsert because neither has higher precedence.
+
+### Available Operations
+
+Use one spelling for each mode: `replace`, `append`, and `upsert`.
+
+| Mode | Behavior at source-precedence and inheritance boundaries |
+| --- | --- |
+| `replace` | Replace the complete lower sequence with the assembled higher sequence. |
+| `append` | Retain lower records and append higher records; duplicate identities fail. |
+| `upsert` | Replace matching records in full, retain unmatched lower records, and append new identities. |
+
+Append and upsert require effective identity. ConfigKit does not infer identity
+from an `id` field or record position. Whole-sequence replacement does not require
+identity agreement with lower YAML policy, but application constraints still apply.
+
+Letters denote record identities; `B*` has the same identity as `B` with a different
+payload:
+
+| Lower | Higher | Mode | Result |
+| --- | --- | --- | --- |
+| `[A, B]` | `[C]` | `append` | `[A, B, C]` |
+| `[A, B]` | `[C]` | `replace` | `[C]` |
+| `[A, B]` | `[B*]` | `append` | Duplicate-identity error |
+| `[A, B]` | `[B*]` | `replace` | `[B*]` |
+| `[A, B]` | `[B*]` | `upsert` | `[A, B*]` |
+| `[A, B]` | `[B*, C]` | `upsert` | `[A, B*, C]` |
+
+Upsert replaces the whole matched record. If lower `B` has `id`, `text`, and
+`temperature`, while higher `B` has only `id` and `text`, the result at this
+boundary has only `id` and `text`. Subsequent defaulting or other lifecycle
+operations retain their own semantics. Mappings continue to combine recursively;
+matching sequence records do not recursively merge fields.
+
+### Mode Selection And Persistence
+
+Select the operation in this order:
+
+1. Reconcile explicit modes within the current tier; conflicting peers fail.
+2. Use the current tier's explicit mode if supplied.
+3. Otherwise retain an applicable explicit mode from a lower tier.
+4. With no explicit mode, infer upsert when effective identity is available.
+5. Otherwise infer replace.
+
+Effective identity can come from an application registration, current declarations,
+or applicable lower policy, subject to the identity agreement and lifetime rules
+below. Determine the inferred mode from available identity before applying the
+operation. A plain higher-tier sequence can therefore use upsert without repeating
+a key declaration.
+
+An explicit mode persists at its applicable destination until a higher-priority
+explicit declaration changes it. New sequence values alone do not reset it, and
+an unrelated overlay leaves it intact. A mode-only declaration can change policy
+without supplying records; it neither clears the collection nor retroactively
+recombines earlier tiers. Later source contributions and inheritance use the
+resulting policy.
+
+| Lower explicit mode | Higher declaration | Effective mode |
+| --- | --- | --- |
+| `append` | Mode omitted | `append` |
+| `append` | `upsert` | `upsert` |
+| `upsert` | `replace` | `replace` |
+| None | Mode omitted, effective identity available | Inferred `upsert` |
+| None | Mode omitted, no effective identity | Inferred `replace` |
+
+Only explicit modes persist. Inferred defaults are recomputed from effective
+identity when needed. Explicit replace continues to apply in subsequent tiers
+until another mode is declared; introducing identity alone does not override it.
+Mode persistence does not preserve YAML identity discarded by replacement, nor
+policies owned by discarded records. Source precedence and inheritance use the
+same destination policy, without separate per-phase settings. CLI assignments
+remain outside mode selection.
+
+### Presence, Empty Sequences, Null, And Kind Mismatches
+
+At a collection's precedence boundary:
+
+| Higher contribution | `replace` | `append` / `upsert` |
+| --- | --- | --- |
+| Absent | Leave the lower value unchanged. | Leave the lower value unchanged. |
+| Explicit `[]` | Clear the sequence. | Contribute no records. |
+| Explicit null or another non-sequence value | Ordinary replacement, subject to application schema. | Error: the operation requires a sequence. |
+
+A declaration alone does not clear a collection. An optional import selecting
+nothing contributes nothing and does not erase existing records. An absent lower
+collection can be initialized by an incoming sequence under any mode, subject to
+identity requirements. Explicit lower null or another non-sequence value is not
+absence: append and upsert fail against it, while replacement remains available.
+An empty incoming sequence bypasses neither kind checks nor identity requirements.
+
+With effective identity and no explicit mode, `prompts: []` uses upsert and does
+not clear the collection. Clearing requires replace, selected in the current tier
+or retained from lower policy. Replacing an identified collection with null also
+requires replace and a schema permitting null. Collection null is distinct from
+a present null identity component inside a record.
+
+### Ordering
+
+Retain existing dependency-ready source scheduling within each tier and authored
+item order within lists. Gathering declarations does not change assembly order;
+identity values do not sort records.
+
+- Append keeps lower records in order, followed by higher records in assembled order.
+- Replace keeps only higher records in their assembled order.
+- Upsert keeps a matched record's lower position and appends new identities in
+  higher-tier order.
+
+For example, lower `[A, B, C]` and higher `[C*, A*, D]` produce `[A*, B, C*, D]`
+under upsert. Whole-sequence replacement gives the higher tier control of the
+complete sequence order.
+
+## Policy Agreement, Lifetime, And Ownership
+
+### Identity Agreement And Replacement
+
+Append and upsert use one agreed identity for all participating records:
+
+| Lower identity | Higher declaration | Result when combining records |
+| --- | --- | --- |
+| `id` | Omitted | Use `id`. |
+| `id` | `id`, `[id]`, or `[id, id]` | Accept the equivalent definition. |
+| `id` | `[namespace, id]` | Error: different definitions. |
+| None | `id` | Establish `id` and validate all participating records. |
+
+Introducing identity for an unkeyed lower sequence requires checking both lower
+and incoming records, including uniqueness within each contribution. With no
+identity from the application or either participating collection, append and
+upsert fail.
+
+YAML identity follows the collection that survives an operation:
+
+| Operation | Effective YAML identity |
+| --- | --- |
+| Append or upsert | Retain or establish the agreed identity. |
+| Whole-sequence replacement | Discard lower YAML identity and use the replacing tier's identity declaration, if any. |
+| Replacement by null or another non-sequence value | Discard lower YAML identity and retain any identity explicitly declared by the replacing tier. |
+
+Application constraints remain authoritative under every operation. Without an
+application constraint, replacement can introduce a different identity, such as
+changing from `id` to `name`. Replacing a sequence without a new key declaration
+or application registration produces an unkeyed sequence. No identity-removal
+directive is needed.
+
+The reset occurs only when the collection value is actually replaced. Declaring
+replace without supplying a replacement does not authorize changing the identity
+of retained records. Explicit mode lifetime is separate from YAML identity
+lifetime.
+
+### Ancestor Replacement
+
+Replacing an enclosing mapping can remove a collection without replacing that
+collection directly. Apply the same identity and ownership rules to this removal:
+discard the removed collection's lower YAML identity, and discard policies owned
+by any removed records. Reintroducing the same path later does not restore those
+discarded policies.
+
+Independently declared explicit modes retain their existing persistence rules.
+Removing a descendant collection does not remove a mode whose declaring scope
+survives. Application registrations remain applicable at matching destinations,
+including destinations reintroduced later. Identity newly declared by the
+replacing tier follows the existing declaration and ownership rules.
+
+For example, a root-level declaration establishes `key: id` and `mode: append`
+for `[catalog, prompts]`. A higher tier replaces `catalog` with null, removing
+the prompts collection and its YAML identity. The independently declared append
+mode survives. If a later tier recreates `catalog.prompts`, append requires
+identity supplied by a new declaration or an application registration. Without
+that identity, loading fails; the later tier can instead explicitly choose
+replace. The replacement of `catalog` itself follows ordinary ancestor-value
+precedence and does not attempt to append null to the removed prompts list.
+
+### Declarations Without Sequence Values
+
+A key declaration without accompanying records applies immediately to an existing
+retained sequence: check agreement and validate its records, without recombining
+earlier tiers. For example, a higher tier that only declares
+`destination: [prompts], key: id` identifies and validates the retained lower
+`prompts` sequence.
+
+If no collection exists, retain the declaration in its owning scope for later
+contributions. An otherwise empty underlay can establish identity for main-tier
+fragments. An explicit identity declaration can also remain pending at a
+non-sequence destination; it does not independently impose a shape requirement.
+The presence and kind rules still determine whether an attempted operation fails.
+Item-local declarations retain their ownership boundaries.
+
+### Lower-Tier Identity During Strict Composition
+
+Resolve applicable identity before composing a higher tier's peers, using
+application registrations, available lower-tier policy, and declarations gathered
+within the higher tier's scope. Authors need not repeat a baseline key solely
+because the higher tier has multiple fragments.
+
+For example, an underlay establishes `key: id` for `prompts`, and two main-tier
+fragments contribute `[foo]` and `[bar]` without a key declaration. They strictly
+compose as `[foo, bar]` before the main tier combines with the accumulated lower
+collection. If both contribute `foo`, composition fails; upsert does not give
+precedence to peers.
+
+Lower YAML identity remains subject to agreement, replacement, and ownership.
+A replacing collection must not be constrained by obsolete lower identity just
+because policy resolution precedes composition. Likewise, a replacement record
+cannot borrow nested identity owned by the discarded lower record, even when
+outer identity or numeric position matches. Application constraints still apply.
+Higher-tier declarations do not retroactively repair invalid lower-tier composition.
+
+### Nested Record Ownership
+
+Attach item-local policies to their owning contributions so they follow surviving
+records independently of numeric indexes. Whole-record upsert discards the old
+record's local policies along with its fields. Whole-sequence replacement discards
+policies owned by removed records. A replacement record supplies its own local
+declarations; equal outer identity or final index does not transfer discarded
+policies.
+
+For example, a source imported as one record into `prompts` can declare:
+
+```yaml
+configkit:
+  sequences:
+    - destination: [steps]
+      key: id
+      mode: append
+id: foo
+steps:
+  - id: prepare
+```
+
+The declaration governs `steps` inside that `foo` record. Retaining `foo` retains
+its local policy, including the explicit mode. Upserting a new `foo` discards the
+old `steps` and its declaration even though the outer record keeps its position.
+
+Policies independently applicable at a destination are not discarded merely
+because an individual record is replaced. Ownership determines which declarations
+are removed. Application registrations remain constraints at their registered
+destinations. Inheritance copies transfer identity and choose operations as
+specified next.
+
+## Sibling Inheritance
+
+Complete source-tier assembly before resolving inheritance. Retain existing
+mapping-sibling scopes and add identified-sequence scopes as described below.
+Both use the existing dependency-ready algorithm. At each inheritance combination,
+the resolved parent's collection is the lower value and the inheriting mapping's
+local collection is the higher value. Mappings combine recursively; upsert within
+an identified sequence remains whole-record replacement. Explicit record
+inheritance is a separate construction step after those replacements.
+
+Use the inheriting destination collection's policy: an applicable explicit mode,
+including one retained through source tiers, wins. Otherwise infer upsert with
+effective identity and replace without it. There is no separate default-append
+rule for inheritance. The same destination policy governs source-precedence and
+inheritance combinations.
+
+YAML identity follows a collection copied through inheritance, subject to agreement
+and replacement rules. A source collection's mode does not automatically become an
+instruction at a different inheriting destination. For example, source `steps`
+has `key: id` and `mode: append`; an inheriting mapping supplies local `steps`
+without an applicable explicit mode. The copied identity selects upsert there.
+An explicit destination append instead selects append. A matching destination key
+agrees; a different key conflicts when combining records, while explicit replace
+can establish new YAML identity subject to application constraints.
+
+For example:
+
+```yaml
+configkit:
+  sequences:
+    - destination: [parent, items]
+      key: id
+child:
+  items:
+    - id: foo
+      value: A
+    - id: bar
+      value: B
+parent:
+  extends: child
+  items:
+    - id: baz
+      value: C
+```
+
+Despite the names, `parent` inherits from `child`. Identity at `[parent, items]`
+selects upsert, producing `[foo, bar, baz]`. If the local sequence instead supplies
+a revised `foo`, the result is `[foo*, bar]`. Append rejects that repeated identity;
+replace retains only the local sequence.
+
+The same ordering, presence, and kind rules apply at inheritance boundaries.
+An omitted local sequence retains the inherited sequence. Explicit `[]` clears it
+under replace and contributes no records under append or upsert. Retain controls
+until inheritance uses them, then exclude them from application decoding.
+
+### Inheritance Between Identified Sequence Records
+
+An identified sequence supplies a sibling scope for its records. Identity can
+come from an application registration or an applicable YAML declaration. A record
+may inherit from one other record in that same sequence using an `extends`
+mapping that selects the parent's complete identity:
+
+```yaml
+configkit:
+  sequences:
+    - destination: [prompts]
+      key: id
+prompts:
+  - id: base
+    model: standard
+    temperature: 0.5
+  - id: concise
+    extends: {id: base}
+    temperature: 0.2
+```
+
+The final `concise` record is:
+
+```yaml
+id: concise
+model: standard
+temperature: 0.2
+```
+
+The `base` record remains in the collection. Parents are ordinary records subject
+to the same decoding and validation as other records; there is no abstract
+template category.
+
+Use mapping-form references for both single and compound keys. With
+`key: [namespace, id]`, a reference can be
+`extends: {namespace: shared, id: base}`. Match components using the collection's
+field-set and scalar-equality rules, including mathematical numeric equality.
+`extends: {id: null}` selects a null-valued identity; `extends: null` cancels a
+relationship.
+
+Parent-reference mappings use ordinary strict mapping composition. Item-local
+helper imports may supply disjoint fields of a compound reference; overlapping
+fields conflict, even when equal. For example, a helper's
+`extends: {namespace: shared}` and the importing record's `extends: {id: base}`
+compose into `extends: {namespace: shared, id: base}`. This assembles one record;
+it does not permit strict peers to contribute duplicate record identities.
+
+Before parent lookup, require the surviving effective reference mapping to contain
+exactly the collection's identity fields, with supported scalar values. Missing
+or extra fields fail. Validate the completed reference during inheritance, rather
+than requiring each helper contribution to be independently complete. Superseded
+references remain historical evidence; earlier strict-composition conflicts
+remain errors.
+
+Every record must supply its complete own identity before inheritance. Preserve
+the child's identity fields and inherit only the parent's nonidentity payload;
+inheritance cannot supply a missing child identity. Child payload fields take
+precedence, mappings combine recursively, and nested sequences use the inheriting
+destination's collection policy. Consume the record's directive before decoding.
+
+Support one parent within the same identified sequence, including forward
+references and chains. Missing parents and cycles, including self-reference,
+fail the load. Cross-collection references, position-based references, multiple
+parents, and inheritance between unidentified sequence members are outside this
+scope. Existing string-form references between mapping siblings retain their
+meaning; they do not become shorthand for sequence identity references.
+
+Resolve relationships against records surviving source-tier assembly and selected
+overlays. A higher layer can replace a parent before inheritance runs; removing
+that parent makes a surviving reference fail. Whole-record upsert discards the
+old record's directive along with its payload. Omitting `extends` in the
+replacement record does not retain the displaced record's relationship. An
+explicit reference to the record's own identity is a cycle, not a way to access
+its lower-tier predecessor.
+
+Reuse Kahn's algorithm with the existing stable canonical semantic-pointer key
+for dependency-ready ordering. Build parent lookup from the sequence's effective
+identity, and preserve the assembled list order independently of resolution
+order. Enclosing relationships resolve before nested relationships. Parent
+templates retain nested instructions for evaluation at each inheriting
+destination, following the existing destination-local inheritance rule. Copied
+collection identity, destination operation selection, and record-owned policy
+transfer follow the rules above.
+
+Record inheritance runs in the existing structural phase, before typed defaults,
+decoding, CLI overrides, and application resolution. Those later stages do not
+trigger another inheritance pass. Preserve inherited presence, origin and
+history evidence, sensitivity, declaring resource contexts, and existing
+expansion budgets just as for mapping-sibling inheritance. Identity lookup must
+not expose identity values in diagnostics. Existing index-based inspection paths
+remain sufficient; no new identity-selector path API is needed.
+
+## Loading Lifecycle
+
+Use the following order:
+
+1. Assemble source tiers and selected overlays, resolving policy before each
+   tier's strict composition and applying collection operations at precedence
+   boundaries.
+2. Resolve sibling inheritance in mappings and identified sequences, then apply
+   typed defaults and decode the assembled configuration.
+3. Apply explicit CLI overrides to that configuration and check supplied values
+   against the application schema.
+4. Run application resolution against the overridden configuration.
+5. Resolve effective secret references, validate the final configuration including
+   application-registered identities, and publish it.
+
+### Typed Defaults And Application Resolution
+
+Source-backed underlays, including application-supplied mapping sources,
+participate in sequence combination. Typed `Definition.Defaults` remains the
+initial decoded state: supplied sequences replace default sequences. Source
+append does not append to a typed default list.
+
+Application resolution retains assignment and recorded-transformation semantics.
+ConfigKit does not automatically combine a callback's assigned sequence with its
+previous value. Application-registered identities receive final validation
+independently of how defaults, CLI overrides, or callbacks produce their values.
+
+### CLI Precedence
+
+CLI overrides are the highest-precedence input to application resolution.
+Supplied sequences replace whole sequences, including sequences inside supplied
+mappings. Mappings retain recursive override behavior; empty sequences and
+permitted nulls retain explicit replacement meaning. Configuration append or
+upsert does not reinterpret CLI assignments. Inheritance and defaulting do not
+run again afterward: overriding assembled `[A, B]` with `[C]` presents `[C]` to
+application resolution without inheritance reintroducing prior records.
+
+Application resolution sees CLI values and their explicit-presence and source
+metadata. `Resolution.Default` preserves explicit CLI values; normalization and
+recorded application transformations may still change them. For example, a derived
+URL can use a CLI hostname, while an explicit CLI URL can be preserved by an
+absence-based default. Retain recording and provenance requirements for those
+transformations.
+
+Reject structural directives such as `extends` and `configkit` in CLI override
+paths or supplied values. CLI input supplies application values after structural
+processing; it cannot supply or cancel inheritance directives, change completed
+relationships, or prevent earlier inheritance failures.
+
+Typed decoding also precedes CLI application, so a CLI replacement cannot repair
+an earlier decoding failure. A source string in an integer field fails before a
+valid CLI integer can apply. This changes current override-before-decoding
+behavior. Semantic validation remains after overrides and application resolution.
+
+Secret lookup uses effective references after overrides and resolution. Neither
+stage bypasses final application identity constraints or semantic validation.
+
+## Validation And Diagnostics
+
+### Structural Identity Checks
+
+Allow item-local helper imports to complete and consume applicable non-main
+source-version headers before checking payload identity for collection assembly.
+A helper can supply an identity component; a consumed header cannot supply one
+after removal.
+
+Structural controls are likewise unavailable as source identity components.
+Check identity against application payload fields, excluding fields acting as
+structural directives even when the engine retains them for later processing.
+For example, `key: extends` with an item containing `extends: null` does not
+supply a null identity: that field cancels inheritance and is consumed. Report
+the ordinary missing-identity-component error at the collection/item path using
+available origins, without exposing values. Do not defer this failure until
+directive consumption or final application validation.
+
+This rule follows a field's structural role; it does not introduce a broader
+restriction on unrelated literal field names. Final application identity checks
+continue to examine the effective typed payload under their existing rules.
+
+Validate each identified contribution before combination and check uniqueness in
+the resulting identified collection, including inheritance operations. Upsert
+matches across precedence are replacements; duplicates within a contribution
+remain errors. Later inheritance, typed defaults, or callbacks cannot repair an
+identity missing at an earlier assembly boundary.
+
+YAML-only identities are enforced during structural composition, precedence
+assembly, and identified inheritance operations. They have no final-result
+guarantee after typed defaulting, CLI overrides, or application resolution. Those
+stages can produce duplicate identities or otherwise violate a YAML-only identity.
+Applications needing a final guarantee register the constraint through the
+application-owned interface.
+
+Adding record inheritance does not expand this guarantee: a YAML declaration
+can enable parent lookup without imposing final validation on values later
+produced by defaults, CLI overrides, or application resolution.
+
+### Final Application Constraints
+
+On successful loading, final effective collections covered by application
+registrations satisfy their declared identity constraints. Retain assembly-time
+checks and validate final effective sequences before publication, after defaults,
+inheritance, overrides, application resolution, and secret lookup.
+
+This covers both exact and dynamic application registrations. Evaluate dynamic
+selectors against the final effective configuration, with independent validation
+for each concrete matched sequence as described under
+[Matching Across The Loading Lifecycle](#matching-across-the-loading-lifecycle).
+
+ConfigKit performs these identity checks; consumers register the constraint
+without needing to repeat duplicate detection in `Definition.Validate`.
+Application-owned semantic validation continues to cover domain-specific rules.
+
+This guarantee covers actual sequences, not path existence or allowed shape.
+It does not require every intermediate callback state to satisfy the constraint
+and does not cover caller mutations after loading returns. Final checks examine
+effective typed values; assembly checks validate authored contributions. A final
+scalar alone does not prove that a source explicitly supplied that field.
+
+Use the existing canonical final-value representation and shared publication
+path as the foundation. The guarantee requires neither a new mutation-history
+system nor exact causal reconstruction for records supplied by defaults or
+callbacks.
+
+### Diagnostics And Inspection
+
+Identity diagnostics identify collection and item paths using available origin
+information without exposing identity values. Conflicting declarations identify
+the collection destination and declaration locations.
+
+For dynamic registrations, use concrete matched collection and item paths for
+value diagnostics, with available origins and existing sensitivity rules.
+Definition-time overlap errors identify the conflicting registrations; they do
+not require or invent a concrete destination from a loaded document.
+
+Correctness includes final values, projected item paths, source destinations,
+provenance, history, sensitivity, and passive resource contexts. Appending can
+change final indexes; replacing changes which contributions survive. Preserve
+the existing loading and inspection contracts through both operations, including
+policy ownership and inheritance transfer.
+
+## Work Needed From The Baseline
+
+The baseline needs the following connected capabilities to reach the target
+state. Execution order and test sequencing belong to the
+[implementation plan](implementation.md):
+
+- **Declarations and public API:** extend application registration to the settled
+  identity model with one public registration type and definition field, an
+  exact-path convenience constructor, and an explicit-selector constructor.
+  Preserve literal `Path` semantics. Parse and consume `configkit.sequences` across
+  supported source forms and selected overlays; preserve placement-relative and
+  item-local scopes. Keep sequence declarations independent of filesystem import
+  enablement.
+- **Identity and policy representation:** implement field-set comparison, typed
+  scalar identity tuples with exact numeric equality, declaration agreement,
+  pending declarations, explicit-mode persistence, replacement lifetime, and
+  record-owned policy transfer. Choose bookkeeping proportional to these semantics.
+  Handle descendant collection removal through ancestor replacement, preserving
+  independently declared modes and application constraints while discarding
+  removed collection identity and record-owned policies.
+- **Registration matching:** support literal and arbitrary-mapping-key segments,
+  detect incompatible selector overlap during definition validation, and resolve
+  applicable constraints at concrete placed destinations. Reuse the identity
+  rules and validator for exact and dynamic registrations. Keep matching separate
+  from source discovery and from YAML declaration ownership.
+- **Assembly handoffs:** make accumulated lower policy available before higher-tier
+  strict composition. Separate source acquisition from policy-aware composition
+  or otherwise revise assembly sequencing; current main-before-underlay composition
+  is not a design constraint. Preserve distinctions among acquisition order,
+  dependency-ready scheduling within tiers, and precedence order. Account for
+  source records, callbacks, main-version checks, and failure diagnostics. Retain
+  version ownership and validation semantics; higher replacement cannot hide
+  lower-tier structural failures.
+- **Combination and inheritance:** implement the three operations with strict peer
+  handling, whole-record replacement, presence rules, stable ordering, and the
+  destination policy rules for inheritance. Extend sibling-scope discovery and
+  parent lookup to identified sequences, validate mapping-form references, and
+  preserve child identities while copying parent payloads. Reuse the shared
+  dependency-ready scheduler and destination-local nested resolution. Preserve
+  provenance, history, sensitivity, paths, and resource contexts as records survive,
+  move, or disappear.
+- **Lifecycle and validation:** move CLI application after inheritance and typed
+  defaults/decoding, preserve explicit-presence and source metadata for application
+  resolution, reject CLI structural directives, and add final application identity
+  validation through the shared publication path. Keep YAML-only enforcement
+  within structural operations. Match dynamic registrations during assembly and
+  inheritance, and against the final effective configuration rather than a saved
+  list of source-backed destinations.
+  Exclude structural controls from source identity checks even while retaining
+  those controls for later inheritance processing.
+- **Documentation and verification:** exercise ordinary and layered loading,
+  imported and inline collections, item-local scopes, version-header consumption,
+  selected overlays, inheritance, typed defaults, CLI, callbacks, and inspection.
+  Cover single and compound parent references, scalar equality, forward references
+  and chains, missing parents and cycles, parent replacement/removal, child identity
+  preservation, nested destination policy, cancellation, and stable output order.
+  Verify helper-composed references, overlapping reference fields, missing or extra
+  fields, and replacement of references together with their containing records.
+  Verify independent namespace uniqueness, root and namespaced registrations,
+  literal punctuation keys, equivalent and incompatible overlaps, and conflicts
+  detected even without a matching input destination. Cover namespaces introduced
+  by imports, placement, higher tiers, inheritance, defaults, CLI, and callbacks;
+  preserve ordinary strict-peer and upsert behavior. Check YAML disagreement under
+  replace, harmless absent or empty collections, schema-owned shape requirements,
+  unrelated sequences remaining unselected, and concrete-path diagnostics.
+  Verify both constructors feed the same registration model, including mixed
+  exact and dynamic registrations, compound identities, and definition-time
+  rejection of empty identity field sets.
+  Cover ancestor replacement followed by collection reintroduction, including a
+  surviving append mode with and without renewed identity, explicit replacement,
+  and application-registered identity. Verify that `extends: null` cannot supply
+  an identity component and that structural-control exclusion uses ordinary
+  value-free missing-identity diagnostics.
+  Update the public contract and consumer examples when behavior lands.
+
+The collection-policy and CLI-ordering decisions are accepted in
+[ADR 0017](../adr/0017-collection-sequence-policy.md) for collection policy and
+[ADR 0018](../adr/0018-cli-after-structural-resolution.md) for CLI ordering.
+Their scoped supersession is recorded in the affected earlier ADRs, and the
+[architecture policy](../policy/architecture.md) reflects the accepted decisions.
+These govern the target architecture; the configuration contract continues to
+describe implemented behavior until the feature lands. This roadmap does not
+select migration machinery.
+
+ADR 0017 now includes inheritance within identified sequences;
+[ADR 0009](../adr/0009-sibling-inheritance.md) records that scope extension in its
+status while preserving its decision content. The architecture policy is aligned.
+Dynamic selectors extend the destinations covered by application registration;
+they retain the accepted application authority and final-validation boundary.
+
+No open feature decisions remain. The implementation plan translates this end
+state into reviewable stages without adding guarantees, restrictions, or
+combination semantics.

example_test.go

@@ -30,6 +30,32 @@ func ExampleNoSource() {
 	// Output: 4 0 <nil>
 }
 
+func ExampleDiff() {
+	type Config struct {
+		Workers int `yaml:"workers"`
+	}
+	before, err := configkit.Load[Config](configkit.Mapping("before", map[string]any{"workers": 2}))
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	after, err := configkit.Load[Config](configkit.Mapping("after", map[string]any{"workers": 4}))
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	differences, err := configkit.Diff(before, after)
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	for _, difference := range differences {
+		fmt.Printf("%s: %s -> %s\n", difference.Path.String(),
+			strings.TrimSpace(difference.Left.YAML), strings.TrimSpace(difference.Right.YAML))
+	}
+	// Output: /workers: 2 -> 4
+}
+
 func ExampleValidation_Error() {
 	type Config struct {
 		First  string `yaml:"first"`

examples/consumers/consumers.go

@@ -29,14 +29,14 @@ type NarratioConfig struct {
 
 // LoadNarratio loads the maintained import fixture with Narratio-shaped rules.
 func LoadNarratio(path string) (*configkit.Result[NarratioConfig], error) {
-	return LoadNarratioProfile(path, "")
+	return LoadNarratioOverlay(path, "")
 }
 
-// LoadNarratioProfile selects a ConfigKit configuration overlay independently
+// LoadNarratioOverlay selects a ConfigKit configuration overlay independently
 // from the application's ModelProfile field.
-func LoadNarratioProfile(path, profile string) (*configkit.Result[NarratioConfig], error) {
+func LoadNarratioOverlay(path, overlay string) (*configkit.Result[NarratioConfig], error) {
 	return configkit.Load(configkit.File(path), configkit.Options[NarratioConfig]{
-		Imports: true, Profile: profile,
+		Imports: true, Overlay: overlay,
 		Definition: configkit.Definition[NarratioConfig]{
 			Defaults: func() NarratioConfig { return NarratioConfig{PollInterval: time.Minute} },
 			Resolve:  resolveNarratio,

examples/consumers/consumers_test.go

@@ -52,6 +52,22 @@ func Example_seriatim() {
 	// Output: visitor true the event occurred <nil> <nil>
 }
 
+func Example_inheritedPresets() {
+	result, err := consumers.LoadPresets(configkit.File(fixture("inheritance", "presets.yml")))
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	path := configkit.Keys("presets", "custom", "enabled")
+	info, _ := result.Metadata().Lookup(path)
+	custom := result.Config.Presets["custom"]
+	fmt.Println(custom.Enabled, result.Metadata().IsExplicit(path), custom.Settings["secondary"]["width"])
+	fmt.Println(info.Origin.Locations[0].SourceID, info.Origin.Inheritance[0].From)
+	// Output:
+	// false true 20
+	// source-1 /presets/base/enabled
+}
+
 func Example_weatherStartup() {
 	const blockedName = "CONFIGKIT_WEATHER_EXAMPLE_BLOCKED"
 	const tokenName = "CONFIGKIT_WEATHER_EXAMPLE_TOKEN"
@@ -92,10 +108,13 @@ func TestMaintainedConsumerFixturesPreserveDomainOutcomes(t *testing.T) {
 	if origin, ok := narratio.Metadata().Lookup(configkit.Keys("artifacts", "transcript", "output")); !ok || origin.Origin.Layer != configkit.LayerResolution {
 		t.Fatalf("artifact origin = %#v, %v", origin, ok)
 	}
-	profiled, err := consumers.LoadNarratioProfile(fixture("narratio", "pipeline.yml"), "testing")
-	if err != nil || profiled.Config.PollInterval.String() != "2m0s" ||
-		profiled.Config.ModelProfile != "concise" || len(profiled.Sources()) != 3 {
-		t.Fatalf("configuration/application profiles = %#v, %v", profiled, err)
+	overlaid, err := consumers.LoadNarratioOverlay(fixture("narratio", "pipeline.yml"), "testing")
+	if err != nil || overlaid.Config.PollInterval.String() != "2m0s" ||
+		overlaid.Config.ModelProfile != "concise" || len(overlaid.Sources()) != 3 {
+		t.Fatalf("configuration overlay and application model profile = %#v, %v", overlaid, err)
+	}
+	if source := overlaid.Sources()[2]; source.Label != "overlays/testing.yml" || source.Layer != configkit.LayerOverlay {
+		t.Fatalf("Narratio overlay source = %#v", source)
 	}
 
 	notarius, err := consumers.LoadNotarius(fixture("notarius", "config.yml"))
@@ -200,7 +219,7 @@ func TestMigrationFixturesFailWithStableCodesAndPaths(t *testing.T) {
 			if len(diagnostics) == 0 || diagnostics[0].Code != test.code || diagnostics[0].Path.String() != test.path {
 				t.Fatalf("diagnostics = %#v", diagnostics)
 			}
-			if strings.Contains(diagnostics[0].Message, "wrong-before-profile") {
+			if strings.Contains(diagnostics[0].Message, "wrong-before-overlay") {
 				t.Fatalf("diagnostic exposed raw parser input: %#v", diagnostics[0])
 			}
 		})

examples/consumers/destinations.go

@@ -0,0 +1,154 @@
+package consumers
+
+import (
+	"errors"
+	"fmt"
+	"io/fs"
+	"path"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+)
+
+// CollectionBundleConfig is a representative application-owned collection
+// shape. Collection and prompt version labels remain ordinary domain fields.
+type CollectionBundleConfig struct {
+	Version     string                          `yaml:"version"`
+	Collections map[string]CollectionDefinition `yaml:"collections"`
+}
+
+type CollectionDefinition struct {
+	Name        string                       `yaml:"name"`
+	Version     string                       `yaml:"version"`
+	Description string                       `yaml:"description"`
+	Prompts     map[string]CollectionPrompt  `yaml:"prompts"`
+	Profiles    map[string]CollectionProfile `yaml:"profiles"`
+}
+
+type CollectionPrompt struct {
+	Version        string              `yaml:"version"`
+	DefaultProfile string              `yaml:"default_profile"`
+	ContentFile    string              `yaml:"content_file"`
+	ResourceText   string              `yaml:"resource_text"`
+	Messages       []CollectionMessage `yaml:"messages"`
+}
+
+type CollectionMessage struct {
+	Role    string `yaml:"role"`
+	Content string `yaml:"content"`
+}
+
+type CollectionProfile struct {
+	Format      string  `yaml:"format"`
+	Temperature float64 `yaml:"temperature"`
+}
+
+// LoadCollectionBundle demonstrates one layered collection load. Its main
+// document owns the application schema version; the mounted entry declares
+// relative destinations for collection metadata, prompt/profile peers, and
+// their nested imports. The operator version override remains a separate tier.
+func LoadCollectionBundle() (*configkit.Result[CollectionBundleConfig], error) {
+	files, err := fs.Sub(bundledFiles, "testdata/collection-bundle")
+	if err != nil {
+		return nil, err
+	}
+	bundle := configkit.NewFilesystem("collection bundle", files)
+	collectionDestination := configkit.Keys("collections", "weatherreporter")
+	promptDestination := collectionDestination.Key("prompts")
+	return configkit.LoadLayered(
+		configkit.Input{Source: configkit.Bytes("application main", []byte("version: 1\n"))},
+		configkit.LayeredOptions[CollectionBundleConfig]{
+			Definition: configkit.Definition[CollectionBundleConfig]{
+				Version: &configkit.VersionRule{Path: configkit.Keys("version"), Check: func(version configkit.Version, _ *configkit.Validation) error {
+					if !version.Present || version.Text != "1" {
+						return errors.New("unsupported application schema version")
+					}
+					return nil
+				}},
+				Resolve:  resolveCollectionResources,
+				Validate: validateCollectionBundle,
+			},
+			Underlays: []configkit.Underlay{
+				{Inputs: []configkit.Input{{
+					Source: bundle.File("collections/weather/collection.yml"), Imports: true, Destination: collectionDestination,
+				}}},
+				{Inputs: []configkit.Input{{
+					Source: bundle.File("operators/weather-prompt-version.yml"), Destination: promptDestination,
+				}}},
+			},
+		},
+	)
+}
+
+func resolveCollectionResources(configuration *CollectionBundleConfig, resolution *configkit.Resolution) error {
+	for collectionID, collection := range configuration.Collections {
+		for promptID, prompt := range collection.Prompts {
+			if prompt.ContentFile == "" {
+				continue
+			}
+			contentPath := configkit.Keys("collections", collectionID, "prompts", promptID, "content_file")
+			info, found := resolution.InputInfo(contentPath)
+			if !found || len(info.Origin.Locations) != 1 {
+				return fmt.Errorf("prompt resource has no single declaring source")
+			}
+			resourceSource, found := resolution.Source(info.Origin.Locations[0].SourceID)
+			if !found {
+				return fmt.Errorf("prompt resource source is unavailable")
+			}
+			context, found := resolution.SourceContext(resourceSource.ID)
+			if !found {
+				return fmt.Errorf("prompt resource context is unavailable")
+			}
+			data, err := fs.ReadFile(context.FS, path.Join(context.BaseDir, prompt.ContentFile))
+			if err != nil {
+				return fmt.Errorf("read prompt resource: %w", err)
+			}
+			prompt.ResourceText = string(data)
+			collection.Prompts[promptID] = prompt
+			resolution.Record(
+				configkit.Keys("collections", collectionID, "prompts", promptID, "resource_text"),
+				configkit.ChangeNormalization, "read source-relative prompt content", contentPath,
+			)
+		}
+		configuration.Collections[collectionID] = collection
+	}
+	return nil
+}
+
+func validateCollectionBundle(configuration *CollectionBundleConfig, validation *configkit.Validation) {
+	for collectionID, collection := range configuration.Collections {
+		for promptID, prompt := range collection.Prompts {
+			if prompt.DefaultProfile == "" {
+				continue
+			}
+			if _, found := collection.Profiles[prompt.DefaultProfile]; !found {
+				validation.Error(configkit.Keys("collections", collectionID, "prompts", promptID, "default_profile"), "must name a profile in this collection")
+			}
+		}
+	}
+}
+
+// ReadCollectionPromptContent follows the effective content field's physical
+// source, including when it was inherited from a peer prompt.
+func ReadCollectionPromptContent(result *configkit.Result[CollectionBundleConfig], collectionID, promptID string) (string, error) {
+	if result == nil {
+		return "", errors.New("collection result is nil")
+	}
+	contentPath := configkit.Keys("collections", collectionID, "prompts", promptID, "content_file")
+	info, found := result.Metadata().Lookup(contentPath)
+	if !found || len(info.Origin.Locations) != 1 {
+		return "", errors.New("prompt resource has no single declaring source")
+	}
+	context, found := result.SourceContext(info.Origin.Locations[0].SourceID)
+	if !found {
+		return "", errors.New("prompt resource context is unavailable")
+	}
+	prompt, found := result.Config.Collections[collectionID].Prompts[promptID]
+	if !found {
+		return "", errors.New("prompt is unavailable")
+	}
+	data, err := fs.ReadFile(context.FS, path.Join(context.BaseDir, prompt.ContentFile))
+	if err != nil {
+		return "", err
+	}
+	return string(data), nil
+}

examples/consumers/destinations_test.go

@@ -0,0 +1,137 @@
+package consumers_test
+
+import (
+	"errors"
+	"fmt"
+	"testing"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+	"gitea.maximumdirect.net/eric/ConfigKit/examples/consumers"
+)
+
+func Example_inputDestination() {
+	mounted, err := configkit.LoadLayered[map[string]any](configkit.Input{
+		Source: configkit.Mapping("weather prompt", map[string]any{
+			"summarize.alerts": map[string]any{"version": "1"},
+		}),
+		Destination: configkit.Keys("collections", "weather.reporter", "prompts"),
+	})
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	root, err := configkit.LoadLayered[map[string]any](configkit.Input{
+		Source: configkit.Bytes("standalone", []byte("name: standalone\n")),
+	})
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	_, found := mounted.Config["collections"]
+	fmt.Println(found, root.Config["name"])
+	// Output: true standalone
+}
+
+func TestCollectionBundleAssemblesMountedPeersAndResources(t *testing.T) {
+	result, err := consumers.LoadCollectionBundle()
+	if err != nil {
+		t.Fatalf("LoadCollectionBundle: %v", err)
+	}
+	collection := result.Config.Collections["weatherreporter"]
+	if result.Config.Version != "1" || collection.Name != "Weather Reporter" || collection.Version != "2.4.0" {
+		t.Fatalf("main and collection versions = %q, %#v", result.Config.Version, collection)
+	}
+	brief := collection.Prompts["brief"]
+	custom := collection.Prompts["custom"]
+	if brief.Version != "2" || brief.DefaultProfile != "compact" || brief.ContentFile != "welcome.txt" || len(brief.Messages) != 1 {
+		t.Fatalf("version-only override replaced lower prompt fields: %#v", brief)
+	}
+	if custom.Version != "2" || custom.DefaultProfile != "weather" || custom.ContentFile != brief.ContentFile ||
+		len(custom.Messages) != 1 || custom.ResourceText != "Welcome to the weather collection.\n" {
+		t.Fatalf("inherited prompt = %#v", custom)
+	}
+	weatherProfile := collection.Profiles["weather"]
+	if weatherProfile.Format != "text" || weatherProfile.Temperature != 0.4 {
+		t.Fatalf("inherited profile = %#v", weatherProfile)
+	}
+	resource, err := consumers.ReadCollectionPromptContent(result, "weatherreporter", "custom")
+	if err != nil || resource != custom.ResourceText {
+		t.Fatalf("inherited prompt resource = %q, %v", resource, err)
+	}
+
+	versionInfo, ok := result.Metadata().Lookup(configkit.Keys("version"))
+	if !ok || !versionInfo.Explicit || versionInfo.Origin.Layer != configkit.LayerBase || len(versionInfo.Origin.Locations) != 1 {
+		t.Fatalf("main-owned version metadata = %#v, %v", versionInfo, ok)
+	}
+	collectionVersionPath := configkit.Keys("collections", "weatherreporter", "version")
+	collectionVersion, ok := result.Metadata().Lookup(collectionVersionPath)
+	if !ok || !collectionVersion.Explicit || collectionVersion.Origin.Layer != configkit.LayerUnderlay || len(collectionVersion.Origin.Locations) != 1 {
+		t.Fatalf("ordinary collection version metadata = %#v, %v", collectionVersion, ok)
+	}
+	var metadataSource configkit.SourceInfo
+	var collectionEntrySource configkit.SourceInfo
+	var versionSource configkit.SourceInfo
+	var parentPromptSource configkit.SourceInfo
+	var promptIndexSource configkit.SourceInfo
+	var profileIndexSource configkit.SourceInfo
+	var promptVersionSource configkit.SourceInfo
+	for _, source := range result.Sources() {
+		switch source.Path {
+		case "collections/weather/collection.yml":
+			collectionEntrySource = source
+		case "collections/weather/metadata.yml":
+			metadataSource = source
+		case "collections/weather/version.yml":
+			versionSource = source
+		case "collections/weather/prompts/index.yml":
+			promptIndexSource = source
+		case "collections/weather/prompts/brief.yml":
+			parentPromptSource = source
+		case "collections/weather/profiles/index.yml":
+			profileIndexSource = source
+		case "operators/weather-prompt-version.yml":
+			promptVersionSource = source
+		}
+	}
+	collectionVersionLocation := collectionVersion.Origin.Locations[0]
+	if collectionEntrySource.ID != "source-2" || collectionEntrySource.Destination.String() != "/collections/weatherreporter" ||
+		metadataSource.ID == "" || metadataSource.Destination.String() != "/collections/weatherreporter" ||
+		metadataSource.ImportedFrom == nil || metadataSource.ImportedFrom.SourceID != collectionEntrySource.ID ||
+		versionSource.ID == "" || collectionVersionLocation.SourceID != versionSource.ID || versionSource.Destination.String() != "/collections/weatherreporter" ||
+		versionSource.ImportedFrom == nil || versionSource.ImportedFrom.SourceID != metadataSource.ID || collectionVersionLocation.Line != 1 {
+		t.Fatalf("nested collection version provenance = entry %#v, metadata %#v, version %#v, location %#v", collectionEntrySource, metadataSource, versionSource, collectionVersionLocation)
+	}
+	if promptIndexSource.ID == "" || promptIndexSource.Destination.String() != "/collections/weatherreporter/prompts" ||
+		promptIndexSource.ImportedFrom == nil || promptIndexSource.ImportedFrom.SourceID != collectionEntrySource.ID ||
+		profileIndexSource.ID == "" || profileIndexSource.Destination.String() != "/collections/weatherreporter/profiles" ||
+		profileIndexSource.ImportedFrom == nil || profileIndexSource.ImportedFrom.SourceID != collectionEntrySource.ID {
+		t.Fatalf("nested prompt/profile placement = prompt %#v profile %#v", promptIndexSource, profileIndexSource)
+	}
+	if promptVersionSource.ID == "" || promptVersionSource.Destination.String() != "/collections/weatherreporter/prompts" || promptVersionSource.UnderlayIndex != 2 {
+		t.Fatalf("prompt version source = %#v", promptVersionSource)
+	}
+	contentPath := configkit.Keys("collections", "weatherreporter", "prompts", "custom", "content_file")
+	contentInfo, ok := result.Metadata().Lookup(contentPath)
+	if !ok || !contentInfo.Explicit || len(contentInfo.Origin.Inheritance) != 1 || len(contentInfo.Origin.Locations) != 1 {
+		t.Fatalf("inherited content field metadata = %#v, %v", contentInfo, ok)
+	}
+	contentLocation := contentInfo.Origin.Locations[0]
+	if parentPromptSource.ID == "" || contentLocation.SourceID != parentPromptSource.ID || contentLocation.Line != 4 ||
+		parentPromptSource.Destination.String() != "/collections/weatherreporter/prompts" {
+		t.Fatalf("inherited resource source coordinates = %#v, source %#v", contentLocation, parentPromptSource)
+	}
+}
+
+func TestCollectionPeerPromptConflictsRemainStrict(t *testing.T) {
+	destination := configkit.Keys("collections", "weatherreporter", "prompts")
+	_, err := configkit.LoadLayered[map[string]any](configkit.Input{Source: configkit.NoSource()}, configkit.LayeredOptions[map[string]any]{
+		Underlays: []configkit.Underlay{{Inputs: []configkit.Input{
+			{Source: configkit.Bytes("peer-a", []byte("brief: {version: '1'}\n")), Destination: destination},
+			{Source: configkit.Bytes("peer-b", []byte("brief: {version: '1'}\n")), Destination: destination},
+		}}},
+	})
+	var loadError *configkit.LoadError
+	if !errors.As(err, &loadError) || len(loadError.Diagnostics()) == 0 || loadError.Diagnostics()[0].Code != configkit.CodeComposeConflict {
+		t.Fatalf("duplicate peer prompt error = %v", err)
+	}
+}

examples/consumers/diff.go

@@ -0,0 +1,136 @@
+package consumers
+
+import (
+	"fmt"
+	"strconv"
+	"strings"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+)
+
+// FileComparisonConfig is a small application-owned shape for comparing
+// caller-selected configuration files.
+type FileComparisonConfig struct {
+	Name    string         `yaml:"name"`
+	Entries map[string]any `yaml:"entries"`
+}
+
+// CompareFiles loads two caller-selected files with the same application
+// definition, then formats only the safe records returned by configkit.Diff.
+func CompareFiles(leftPath, rightPath string) (string, error) {
+	options := configkit.Options[FileComparisonConfig]{
+		Definition: configkit.Definition[FileComparisonConfig]{
+			Defaults: func() FileComparisonConfig {
+				return FileComparisonConfig{Entries: map[string]any{}}
+			},
+		},
+	}
+	left, err := configkit.Load(configkit.File(leftPath), options)
+	if err != nil {
+		return "", fmt.Errorf("load left configuration: %w", err)
+	}
+	right, err := configkit.Load(configkit.File(rightPath), options)
+	if err != nil {
+		return "", fmt.Errorf("load right configuration: %w", err)
+	}
+	differences, err := configkit.Diff(left, right)
+	if err != nil {
+		return "", err
+	}
+	return FormatDifferences(differences), nil
+}
+
+// RolloutConfig models a service endpoint whose URL is derived by the
+// application resolver and whose credential reference is classified.
+type RolloutConfig struct {
+	Service RolloutService `yaml:"service"`
+}
+
+// RolloutService contains endpoint settings and a classified credential.
+type RolloutService struct {
+	Endpoint   RolloutEndpoint   `yaml:"endpoint"`
+	Credential RolloutCredential `yaml:"credential"`
+}
+
+// RolloutEndpoint is the effective destination used by the application.
+type RolloutEndpoint struct {
+	Host string `yaml:"host"`
+	Port int    `yaml:"port"`
+	URL  string `yaml:"url"`
+}
+
+// RolloutCredential holds only a typed secret reference in configuration.
+type RolloutCredential struct {
+	Token configkit.Secret `yaml:"token" configkit:"secret"`
+}
+
+// LoadRollout loads a base file, optionally with one selected configuration
+// overlay. Resolution derives the effective URL after the overlay is applied.
+func LoadRollout(basePath, overlayName string) (*configkit.Result[RolloutConfig], error) {
+	return configkit.Load(configkit.File(basePath), configkit.Options[RolloutConfig]{
+		Overlay: overlayName,
+		Definition: configkit.Definition[RolloutConfig]{
+			Resolve: resolveRollout,
+		},
+	})
+}
+
+// CompareRolloutOverlay compares the effective base configuration with that
+// same file plus one selected overlay. It presents only safe diff records.
+func CompareRolloutOverlay(basePath, overlayName string) (string, error) {
+	base, err := LoadRollout(basePath, "")
+	if err != nil {
+		return "", fmt.Errorf("load base configuration: %w", err)
+	}
+	withOverlay, err := LoadRollout(basePath, overlayName)
+	if err != nil {
+		return "", fmt.Errorf("load overlay configuration: %w", err)
+	}
+	differences, err := configkit.Diff(base, withOverlay)
+	if err != nil {
+		return "", err
+	}
+	return FormatDifferences(differences), nil
+}
+
+func resolveRollout(configuration *RolloutConfig, resolution *configkit.Resolution) error {
+	endpoint := &configuration.Service.Endpoint
+	if endpoint.Host == "" || endpoint.Port <= 0 {
+		return nil
+	}
+	url := fmt.Sprintf("https://%s:%d", endpoint.Host, endpoint.Port)
+	if endpoint.URL != url {
+		endpoint.URL = url
+		resolution.Record(configkit.Keys("service", "endpoint", "url"),
+			configkit.ChangeNormalization, "derived endpoint URL",
+			configkit.Keys("service", "endpoint", "host"),
+			configkit.Keys("service", "endpoint", "port"))
+	}
+	return nil
+}
+
+// FormatDifferences is consumer-owned presentation. It distinguishes absence
+// through Present and reports withheld changes through HiddenChanged without
+// interpreting redaction marker text.
+func FormatDifferences(differences []configkit.Difference) string {
+	if len(differences) == 0 {
+		return "No effective changes.\n"
+	}
+	var output strings.Builder
+	for _, difference := range differences {
+		fmt.Fprintf(&output, "%s: left=%s right=%s", difference.Path.String(),
+			formatSnapshot(difference.Left), formatSnapshot(difference.Right))
+		if difference.HiddenChanged {
+			output.WriteString(" hidden-change")
+		}
+		output.WriteByte('\n')
+	}
+	return output.String()
+}
+
+func formatSnapshot(snapshot configkit.ValueSnapshot) string {
+	if !snapshot.Present {
+		return "<absent>"
+	}
+	return strconv.Quote(snapshot.YAML)
+}

examples/consumers/diff_test.go

@@ -0,0 +1,187 @@
+package consumers_test
+
+import (
+	"fmt"
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+	"gitea.maximumdirect.net/eric/ConfigKit/examples/consumers"
+)
+
+func ExampleCompareFiles() {
+	directory, err := os.MkdirTemp("", "configkit-diff-files-")
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	defer os.RemoveAll(directory)
+	leftPath := filepath.Join(directory, "before.yml")
+	rightPath := filepath.Join(directory, "after.yml")
+	if err := os.WriteFile(leftPath, []byte("name: before\nentries: {}\n"), 0o600); err != nil {
+		fmt.Println(err)
+		return
+	}
+	if err := os.WriteFile(rightPath, []byte("name: after\nentries:\n  added: null\n"), 0o600); err != nil {
+		fmt.Println(err)
+		return
+	}
+	output, err := consumers.CompareFiles(leftPath, rightPath)
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	fmt.Print(output)
+	// Output:
+	// /entries/added: left=<absent> right="null\n"
+	// /name: left="before\n" right="after\n"
+}
+
+func ExampleCompareRolloutOverlay() {
+	const baseToken = "CONFIGKIT_DIFF_EXAMPLE_BASE_TOKEN"
+	const overlayToken = "CONFIGKIT_DIFF_EXAMPLE_OVERLAY_TOKEN"
+	defer preserveEnvironment(baseToken)()
+	defer preserveEnvironment(overlayToken)()
+	_ = os.Setenv(baseToken, "disposable-example-base-credential")
+	_ = os.Setenv(overlayToken, "disposable-example-overlay-credential")
+
+	directory, err := os.MkdirTemp("", "configkit-diff-overlay-")
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	defer os.RemoveAll(directory)
+	basePath := filepath.Join(directory, "config.yml")
+	overlayPath := filepath.Join(directory, "overlays", "canary.yml")
+	if err := os.MkdirAll(filepath.Dir(overlayPath), 0o700); err != nil {
+		fmt.Println(err)
+		return
+	}
+	base := "service:\n  endpoint:\n    host: base.example\n    port: 8443\n  credential:\n    token: " + baseToken + "\n"
+	overlay := "service:\n  endpoint:\n    host: canary.example\n  credential:\n    token: " + overlayToken + "\n"
+	if err := os.WriteFile(basePath, []byte(base), 0o600); err != nil {
+		fmt.Println(err)
+		return
+	}
+	if err := os.WriteFile(overlayPath, []byte(overlay), 0o600); err != nil {
+		fmt.Println(err)
+		return
+	}
+	output, err := consumers.CompareRolloutOverlay(basePath, "canary")
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	fmt.Print(output)
+	// Output:
+	// /service/credential/token: left="'[REDACTED]'\n" right="'[REDACTED]'\n" hidden-change
+	// /service/endpoint/host: left="base.example\n" right="canary.example\n"
+	// /service/endpoint/url: left="https://base.example:8443\n" right="https://canary.example:8443\n"
+}
+
+func TestCompareFilesUsesCallerSelectedSourcesAndSafePresence(t *testing.T) {
+	directory := t.TempDir()
+	leftPath := filepath.Join(directory, "before.yml")
+	rightPath := filepath.Join(directory, "after.yml")
+	writeConfiguration(t, leftPath, "name: before\nentries: {}\n")
+	writeConfiguration(t, rightPath, "name: after\nentries:\n  added: null\n")
+
+	output, err := consumers.CompareFiles(leftPath, rightPath)
+	if err != nil {
+		t.Fatal(err)
+	}
+	want := "/entries/added: left=<absent> right=\"null\\n\"\n" +
+		"/name: left=\"before\\n\" right=\"after\\n\"\n"
+	if output != want {
+		t.Fatalf("comparison output = %q, want %q", output, want)
+	}
+	if again, err := consumers.CompareFiles(leftPath, rightPath); err != nil || again != output {
+		t.Fatalf("repeat output = %q, %v", again, err)
+	}
+}
+
+func TestCompareFilesReturnsEachLoadError(t *testing.T) {
+	directory := t.TempDir()
+	leftPath := filepath.Join(directory, "before.yml")
+	rightPath := filepath.Join(directory, "after.yml")
+	writeConfiguration(t, leftPath, "name: before\n")
+	writeConfiguration(t, rightPath, "unknown: invalid\n")
+
+	output, err := consumers.CompareFiles(leftPath, rightPath)
+	if err == nil || output != "" || !strings.Contains(err.Error(), "load right configuration") {
+		t.Fatalf("right load failure = %q, %v", output, err)
+	}
+	if _, err := consumers.CompareFiles(filepath.Join(directory, "missing.yml"), rightPath); err == nil || !strings.Contains(err.Error(), "load left configuration") {
+		t.Fatalf("left load failure = %v", err)
+	}
+}
+
+func TestCompareRolloutOverlayShowsResolvedSafeChangesAndRetainsSiblings(t *testing.T) {
+	t.Setenv("CONFIGKIT_DIFF_BASE_TOKEN", "disposable-base-credential")
+	t.Setenv("CONFIGKIT_DIFF_OVERLAY_TOKEN", "disposable-overlay-credential")
+	directory := t.TempDir()
+	basePath := filepath.Join(directory, "config.yml")
+	overlayPath := filepath.Join(directory, "overlays", "canary.yml")
+	if err := os.MkdirAll(filepath.Dir(overlayPath), 0o700); err != nil {
+		t.Fatal(err)
+	}
+	writeConfiguration(t, basePath, "service:\n  endpoint:\n    host: base.example\n    port: 8443\n  credential:\n    token: CONFIGKIT_DIFF_BASE_TOKEN\n")
+	writeConfiguration(t, overlayPath, "service:\n  endpoint:\n    host: canary.example\n  credential:\n    token: CONFIGKIT_DIFF_OVERLAY_TOKEN\n")
+
+	base, err := consumers.LoadRollout(basePath, "")
+	if err != nil {
+		t.Fatal(err)
+	}
+	withOverlay, err := consumers.LoadRollout(basePath, "canary")
+	if err != nil {
+		t.Fatal(err)
+	}
+	if base.Config.Service.Endpoint.Port != 8443 || withOverlay.Config.Service.Endpoint.Port != 8443 {
+		t.Fatalf("lower-priority port was not retained: base=%#v overlay=%#v", base.Config.Service.Endpoint, withOverlay.Config.Service.Endpoint)
+	}
+	urlInfo, found := withOverlay.Metadata().Lookup(configkit.Keys("service", "endpoint", "url"))
+	if !found || urlInfo.Origin.Layer != configkit.LayerResolution || withOverlay.Config.Service.Endpoint.URL != "https://canary.example:8443" {
+		t.Fatalf("resolved endpoint = %#v, metadata=%#v, %v", withOverlay.Config.Service.Endpoint, urlInfo, found)
+	}
+
+	differences, err := configkit.Diff(base, withOverlay)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(differences) != 3 {
+		t.Fatalf("differences = %#v", differences)
+	}
+	wantPaths := []string{"/service/credential/token", "/service/endpoint/host", "/service/endpoint/url"}
+	for index, difference := range differences {
+		if got := difference.Path.String(); got != wantPaths[index] {
+			t.Errorf("path[%d] = %q, want %q", index, got, wantPaths[index])
+		}
+	}
+	secret := differences[0]
+	if !secret.Left.Redacted || !secret.Right.Redacted || !secret.HiddenChanged || secret.Left.YAML != "'[REDACTED]'\n" || secret.Right.YAML != "'[REDACTED]'\n" {
+		t.Fatalf("classified credential difference = %#v", secret)
+	}
+	if strings.Contains(consumers.FormatDifferences(differences), "disposable-") {
+		t.Fatal("formatted differences exposed credential contents")
+	}
+
+	output, err := consumers.CompareRolloutOverlay(basePath, "canary")
+	if err != nil || output != consumers.FormatDifferences(differences) {
+		t.Fatalf("formatted base/overlay comparison = %q, %v", output, err)
+	}
+}
+
+func TestFormatDifferencesUsesRecordMembershipForNoChange(t *testing.T) {
+	if got := consumers.FormatDifferences(nil); got != "No effective changes.\n" {
+		t.Fatalf("no-change output = %q", got)
+	}
+}
+
+func writeConfiguration(t *testing.T, filename, contents string) {
+	t.Helper()
+	if err := os.WriteFile(filename, []byte(contents), 0o600); err != nil {
+		t.Fatal(err)
+	}
+}

examples/consumers/inheritance.go

@@ -0,0 +1,20 @@
+package consumers
+
+import configkit "gitea.maximumdirect.net/eric/ConfigKit"
+
+// PresetCatalog is ordinary application data; no inheritance fields or
+// registration hooks are required in the typed configuration.
+type PresetCatalog struct {
+	Presets map[string]Preset `yaml:"presets"`
+}
+
+type Preset struct {
+	Enabled  bool                      `yaml:"enabled"`
+	Settings map[string]map[string]int `yaml:"settings"`
+}
+
+// LoadPresets performs one load; callers may inspect inherited presence and
+// source origins on the returned result before choosing a preset to execute.
+func LoadPresets(source configkit.Source) (*configkit.Result[PresetCatalog], error) {
+	return configkit.Load[PresetCatalog](source)
+}

examples/consumers/inheritance_test.go

@@ -0,0 +1,35 @@
+package consumers_test
+
+import (
+	"testing"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+	"gitea.maximumdirect.net/eric/ConfigKit/examples/consumers"
+)
+
+func TestPresetCatalogOneLoadAndDerivation(t *testing.T) {
+	result, err := consumers.LoadPresets(configkit.File(fixture("inheritance", "presets.yml")))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if result.Config.Presets["base"].Settings["secondary"]["width"] != 10 ||
+		result.Config.Presets["custom"].Settings["secondary"]["width"] != 20 ||
+		result.Config.Presets["custom"].Enabled {
+		t.Fatalf("expanded presets = %#v", result.Config.Presets)
+	}
+	enabled := configkit.Keys("presets", "custom", "enabled")
+	if !result.Metadata().IsExplicit(enabled) {
+		t.Fatal("inherited false lost explicit presence")
+	}
+	info, found := result.Metadata().Lookup(enabled)
+	if !found || info.Origin.Locations[0].SourceID != "source-1" ||
+		len(info.Origin.Inheritance) != 1 || info.Origin.Inheritance[0].From.String() != "/presets/base/enabled" {
+		t.Fatalf("origin and derivation = %#v, %v", info, found)
+	}
+	nested := configkit.Keys("presets", "custom", "settings", "secondary", "width")
+	info, found = result.Metadata().Lookup(nested)
+	if !found || len(info.Origin.Inheritance) != 1 ||
+		info.Origin.Inheritance[0].From.String() != "/presets/custom/settings/primary/width" {
+		t.Fatalf("destination-local derivation = %#v, %v", info, found)
+	}
+}

examples/consumers/layered.go

@@ -0,0 +1,119 @@
+package consumers
+
+import (
+	"embed"
+	"errors"
+	"io/fs"
+	"path"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+)
+
+//go:embed testdata/native-source testdata/layered-bundle testdata/collection-bundle testdata/ordinary-imports testdata/source-version
+var bundledFiles embed.FS
+
+// BundledConfig is a small configuration loaded directly from an embedded file.
+type BundledConfig struct {
+	Name    string `yaml:"name"`
+	Enabled bool   `yaml:"enabled"`
+}
+
+type PromptBundleConfig struct {
+	Name        string                  `yaml:"name"`
+	Version     string                  `yaml:"version"`
+	Description string                  `yaml:"description"`
+	Prompts     map[string]BundlePrompt `yaml:"prompts"`
+}
+
+type BundlePrompt struct {
+	Summary string `yaml:"summary"`
+}
+
+// LoadPromptBundle demonstrates ordinary Load with opt-in imports whose
+// destinations are declared relative to the entry document.
+func LoadPromptBundle() (*configkit.Result[PromptBundleConfig], error) {
+	files, err := fs.Sub(bundledFiles, "testdata/ordinary-imports")
+	if err != nil {
+		return nil, err
+	}
+	return configkit.Load[PromptBundleConfig](
+		configkit.NewFilesystem("prompt bundle", files).File("config.yml"),
+		configkit.Options[PromptBundleConfig]{Imports: true},
+	)
+}
+
+// LoadBundled demonstrates a native filesystem source without layered options.
+func LoadBundled() (*configkit.Result[BundledConfig], error) {
+	files, err := fs.Sub(bundledFiles, "testdata/native-source")
+	if err != nil {
+		return nil, err
+	}
+	return configkit.Load[BundledConfig](configkit.NewFilesystem("bundled", files).File("config.yml"))
+}
+
+// CatalogConfig keeps resource names and domain rules in the application.
+type CatalogConfig struct {
+	Mode      string            `yaml:"mode"`
+	Enabled   bool              `yaml:"enabled"`
+	Workers   int               `yaml:"workers"`
+	Templates map[string]string `yaml:"templates"`
+	Labels    map[string]string `yaml:"labels"`
+}
+
+// LoadCatalog combines two underlay tiers beneath an independently located
+// host main file. ConfigKit assembles and validates the complete result once.
+func LoadCatalog(mainPath string) (*configkit.Result[CatalogConfig], error) {
+	files, err := fs.Sub(bundledFiles, "testdata/layered-bundle")
+	if err != nil {
+		return nil, err
+	}
+	bundle := configkit.NewFilesystem("catalog bundle", files)
+	return configkit.LoadLayered(configkit.Input{Source: configkit.File(mainPath)},
+		configkit.LayeredOptions[CatalogConfig]{
+			Underlays: []configkit.Underlay{
+				{Inputs: []configkit.Input{{Source: bundle.File("defaults/config.yml"), Imports: true}}},
+				{Inputs: []configkit.Input{{Source: configkit.Mapping("local policy", map[string]any{
+					"workers": 2,
+					"labels":  map[string]any{"tenant": "example"},
+				})}}},
+			},
+			Definition: configkit.Definition[CatalogConfig]{
+				Defaults: func() CatalogConfig { return CatalogConfig{Mode: "demo"} },
+				Validate: validateCatalog,
+			},
+		})
+}
+
+func validateCatalog(configuration *CatalogConfig, validation *configkit.Validation) {
+	if configuration.Mode != "demo" && configuration.Mode != "production" {
+		validation.Error(configkit.Keys("mode"), "must be demo or production")
+	}
+	if configuration.Workers < 0 {
+		validation.Error(configkit.Keys("workers"), "must be nonnegative")
+	}
+	if name := configuration.Templates["welcome"]; name == "." || !fs.ValidPath(name) {
+		validation.Error(configkit.Keys("templates", "welcome"), "must be a relative resource name")
+	}
+}
+
+// ReadWelcome follows the effective template field's one declaring source to
+// its retained resource context. Opening and interpreting the resource remain
+// application work; this does not reopen configuration or change metadata.
+func ReadWelcome(result *configkit.Result[CatalogConfig]) (string, error) {
+	if result == nil {
+		return "", errors.New("catalog result is nil")
+	}
+	info, found := result.Metadata().Lookup(configkit.Keys("templates", "welcome"))
+	if !found || len(info.Origin.Locations) != 1 {
+		return "", errors.New("welcome template has no single declaring source")
+	}
+	context, found := result.SourceContext(info.Origin.Locations[0].SourceID)
+	if !found {
+		return "", errors.New("welcome template has no resource context")
+	}
+	data, err := fs.ReadFile(context.FS, path.Join(context.BaseDir, result.Config.Templates["welcome"]))
+	if err != nil {
+		return "", err
+	}
+	return string(data), nil
+}

examples/consumers/layered_test.go

@@ -0,0 +1,151 @@
+package consumers_test
+
+import (
+	"errors"
+	"fmt"
+	"strings"
+	"testing"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+	"gitea.maximumdirect.net/eric/ConfigKit/examples/consumers"
+)
+
+func Example_nativeFilesystem() {
+	result, err := consumers.LoadBundled()
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	fmt.Println(result.Config.Name, result.Config.Enabled)
+	// Output: embedded false
+}
+
+func Example_rootRelativeImports() {
+	result, err := consumers.LoadPromptBundle()
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	source := result.Sources()[2]
+	fmt.Println(result.Config.Name, result.Config.Version, result.Config.Prompts["brief"].Summary, source.Destination.String(), err)
+	// Output: Root import example 1.0 A root-loaded prompt. /prompts <nil>
+}
+
+func TestPromptBundleUsesRelativeImportDestinations(t *testing.T) {
+	result, err := consumers.LoadPromptBundle()
+	if err != nil {
+		t.Fatal(err)
+	}
+	if result.Config.Name != "Root import example" || result.Config.Version != "1.0" || result.Config.Prompts["brief"].Summary != "A root-loaded prompt." {
+		t.Fatalf("root import result = %#v", result.Config)
+	}
+	sources := result.Sources()
+	if len(sources) != 3 || sources[0].Destination.String() != configkit.Keys().String() ||
+		sources[1].Path != "metadata.yml" || sources[1].Destination.String() != configkit.Keys().String() ||
+		sources[2].Path != "prompts/brief.yml" || sources[2].Destination.String() != "/prompts" ||
+		sources[2].ImportedFrom == nil || sources[2].ImportedFrom.SourceID != sources[0].ID {
+		t.Fatalf("root import sources = %#v", sources)
+	}
+}
+
+func Example_layeredCatalog() {
+	result, err := consumers.LoadCatalog(fixture("layered-main", "config.yml"))
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	welcome, err := consumers.ReadWelcome(result)
+	fmt.Println(result.Config.Mode, result.Config.Enabled, result.Config.Workers,
+		result.Config.Labels["origin"], result.Config.Labels["tenant"], strings.TrimSpace(welcome), err)
+	// Output: production false 0 bundle example Welcome from the bundle. <nil>
+}
+
+func Example_sourceVersionAdoption() {
+	result, err := consumers.LoadVersionedService(configkit.SourceVersionModeIfPresent)
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	checks, redacted := 0, true
+	for _, source := range result.Sources() {
+		if check := source.VersionCheck; check != nil {
+			checks++
+			redacted = redacted && check.Redacted && check.Version.Text == ""
+		}
+	}
+	required, requiredErr := consumers.LoadVersionedService(configkit.SourceVersionModeRequired)
+	fmt.Println(result.Config.Service.Version, result.Config.Service.Metadata.Release,
+		checks, redacted, required != nil && requiredErr == nil)
+	// Output: service-v2 application-2026.10 2 true true
+}
+
+func TestVersionedServiceModesAndMountedProtection(t *testing.T) {
+	for _, mode := range []configkit.SourceVersionMode{
+		configkit.SourceVersionModeIfPresent,
+		configkit.SourceVersionModeRequired,
+	} {
+		result, err := consumers.LoadVersionedService(mode)
+		if err != nil || result == nil || result.Config.Service.Version != "service-v2" ||
+			result.Config.Service.Name != "example service" || result.Config.Service.Endpoint != "https://service.example.invalid" ||
+			result.Config.Service.Metadata.Release != "application-2026.10" {
+			t.Fatalf("versioned service mode %d = %#v, %v", mode, result, err)
+		}
+		sources := result.Sources()
+		if len(sources) != 2 {
+			t.Fatalf("versioned service sources = %#v", sources)
+		}
+		for _, source := range sources {
+			check := source.VersionCheck
+			if check == nil || check.Path.String() != "/version" || !check.Version.Present ||
+				!check.Redacted || check.Version.Text != "" {
+				t.Fatalf("safe source version evidence = %#v", sources)
+			}
+		}
+	}
+	_, err := consumers.LoadVersionedService(configkit.SourceVersionModeIfPresent,
+		configkit.Override{Path: configkit.Keys("service", "version"), Value: "cli-v3", Flag: "--service-version"})
+	var loadError *configkit.LoadError
+	if !errors.As(err, &loadError) || len(loadError.Diagnostics()) != 1 ||
+		loadError.Diagnostics()[0].Code != configkit.CodeOverrideVersion ||
+		loadError.Diagnostics()[0].Path.String() != "/service/version" {
+		t.Fatalf("mounted main version override error = %v", err)
+	}
+}
+
+func TestCatalogRetainsPresenceAndSourceContext(t *testing.T) {
+	result, err := consumers.LoadCatalog(fixture("layered-main", "config.yml"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	for _, key := range []string{"enabled", "workers"} {
+		info, found := result.Metadata().Lookup(configkit.Keys(key))
+		if !found || !info.Explicit || info.Origin.Layer != configkit.LayerBase {
+			t.Fatalf("%s presence = %#v, %v", key, info, found)
+		}
+	}
+	info, found := result.Metadata().Lookup(configkit.Keys("templates", "welcome"))
+	if !found || info.Origin.Layer != configkit.LayerUnderlay || len(info.Origin.Locations) != 1 {
+		t.Fatalf("welcome origin = %#v, %v", info, found)
+	}
+	context, found := result.SourceContext(info.Origin.Locations[0].SourceID)
+	if !found || context.BaseDir != "defaults" {
+		t.Fatalf("welcome context = %#v, %v", context, found)
+	}
+	welcome, err := consumers.ReadWelcome(result)
+	if err != nil || strings.TrimSpace(welcome) != "Welcome from the bundle." {
+		t.Fatalf("welcome = %q, %v", welcome, err)
+	}
+	if result.Config.Labels["tenant"] != "example" || result.Config.Labels["origin"] != "bundle" {
+		t.Fatalf("recursive underlay labels = %#v", result.Config.Labels)
+	}
+}
+
+func TestCatalogValidationRejectsInvalidMode(t *testing.T) {
+	_, err := consumers.LoadCatalog(fixture("layered-main", "invalid.yml"))
+	var loadError *configkit.LoadError
+	if !errors.As(err, &loadError) || len(loadError.Diagnostics()) == 0 ||
+		loadError.Diagnostics()[0].Code != configkit.CodeValidateError ||
+		loadError.Diagnostics()[0].Path.String() != "/mode" {
+		t.Fatalf("invalid mode error = %v", err)
+	}
+}

examples/consumers/prompt_catalog.go

@@ -0,0 +1,68 @@
+package consumers
+
+import (
+	"strings"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+)
+
+// PromptCatalogConfig is an application-owned prompt collection shape.
+type PromptCatalogConfig struct {
+	Prompts []PromptRecord `yaml:"prompts"`
+}
+
+// PromptRecord keeps its identity and message meaning in the consumer.
+type PromptRecord struct {
+	ID       string   `yaml:"id"`
+	Name     string   `yaml:"name"`
+	Messages []string `yaml:"messages"`
+	Resource string   `yaml:"resource"`
+}
+
+// LoadPromptCatalog discovers explicitly registered prompt records in the main
+// tier and replaces the lower-priority default collection as a whole.
+func LoadPromptCatalog(mainPath string) (*configkit.Result[PromptCatalogConfig], error) {
+	return configkit.LoadLayered(configkit.Input{
+		Source:  configkit.File(mainPath),
+		Imports: true,
+	}, configkit.LayeredOptions[PromptCatalogConfig]{
+		Underlays: []configkit.Underlay{{Inputs: []configkit.Input{{
+			Source: configkit.Bytes("default prompts", []byte("prompts:\n  - id: default\n    name: Default prompt\n    messages: [Default message]\n")),
+		}}}},
+		ImportDefaults: configkit.ImportDefaults{
+			Pattern: []string{"curated*.yml", "generated-*.yml"},
+		},
+		Definition: promptCatalogDefinition(),
+	})
+}
+
+// LoadRequiredPromptCatalog keeps the common one-file import shape small.
+func LoadRequiredPromptCatalog(mainPath string) (*configkit.Result[PromptCatalogConfig], error) {
+	return configkit.Load(configkit.File(mainPath), configkit.Options[PromptCatalogConfig]{
+		Imports: true, Definition: promptCatalogDefinition(),
+	})
+}
+
+func promptCatalogDefinition() configkit.Definition[PromptCatalogConfig] {
+	return configkit.Definition[PromptCatalogConfig]{
+		SequenceCompositions: []configkit.SequenceComposition{{
+			Path:          configkit.Keys("prompts"),
+			IdentityField: "id",
+		}},
+		Validate: validatePromptCatalog,
+	}
+}
+
+func validatePromptCatalog(configuration *PromptCatalogConfig, validation *configkit.Validation) {
+	for index, prompt := range configuration.Prompts {
+		if strings.TrimSpace(prompt.Name) == "" {
+			validation.Error(configkit.Keys("prompts").Index(index).Key("name"), "must not be empty")
+		}
+		if len(prompt.Messages) == 0 {
+			validation.Error(configkit.Keys("prompts").Index(index).Key("messages"), "must contain at least one message")
+		}
+	}
+	if len(configuration.Prompts) == 0 {
+		validation.Error(configkit.Keys("prompts"), "must contain at least one prompt")
+	}
+}

examples/consumers/prompt_catalog_test.go

@@ -0,0 +1,135 @@
+package consumers_test
+
+import (
+	"errors"
+	"fmt"
+	"io/fs"
+	pathpkg "path"
+	"path/filepath"
+	"testing"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+	"gitea.maximumdirect.net/eric/ConfigKit/examples/consumers"
+)
+
+func Example_identifiedPromptCatalog() {
+	result, err := consumers.LoadPromptCatalog(fixture("prompt-catalog", "config.yml"))
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	ids := make([]string, 0, len(result.Config.Prompts))
+	for _, prompt := range result.Config.Prompts {
+		ids = append(ids, prompt.ID)
+	}
+	emptySearches := 0
+	for _, source := range result.Sources() {
+		for _, selection := range source.ImportSelections {
+			if selection.Selector == configkit.ImportSelectorDirectory && selection.Outcome == configkit.ImportSelectionNone {
+				emptySearches++
+			}
+		}
+	}
+	fmt.Println(ids, emptySearches, err)
+	// Output: [weather curated welcome main] 1 <nil>
+}
+
+func Example_requiredPromptCatalog() {
+	result, err := consumers.LoadRequiredPromptCatalog(fixture("prompt-catalog", "simple.yml"))
+	if err != nil {
+		fmt.Println(err)
+		return
+	}
+	fmt.Println(result.Config.Prompts[0].ID, err)
+	// Output: required <nil>
+}
+
+func TestPromptCatalogImportsRegisteredItemsAndReplacesDefaults(t *testing.T) {
+	result, err := consumers.LoadPromptCatalog(fixture("prompt-catalog", "config.yml"))
+	if err != nil {
+		var loadError *configkit.LoadError
+		if errors.As(err, &loadError) {
+			t.Fatalf("load prompt catalog: %v (%#v)", err, loadError.Diagnostics())
+		}
+		t.Fatal(err)
+	}
+	if len(result.Config.Prompts) != 4 {
+		t.Fatalf("prompt records = %#v", result.Config.Prompts)
+	}
+	ids := []string{"weather", "curated", "welcome", "main"}
+	for index, id := range ids {
+		if result.Config.Prompts[index].ID != id {
+			t.Fatalf("prompt %d ID = %q, want %q", index, result.Config.Prompts[index].ID, id)
+		}
+	}
+	for _, prompt := range result.Config.Prompts {
+		if prompt.ID == "default" {
+			t.Fatalf("lower-priority default survived main sequence replacement: %#v", result.Config.Prompts)
+		}
+		if prompt.ID == "ignored" {
+			t.Fatalf("excluded generated prompt was selected: %#v", result.Config.Prompts)
+		}
+	}
+
+	var itemImports, helperImports int
+	for _, source := range result.Sources() {
+		switch filepath.Base(source.Path) {
+		case "curated.yml", "weather-feature.yml", "generated-welcome.yml":
+			if source.PlacementMode != configkit.PlacementModeSequenceItem || source.Destination.String() != "/prompts" {
+				t.Fatalf("item source placement = %#v", source)
+			}
+			itemImports++
+		case "curated-id.yml":
+			if source.PlacementMode != configkit.PlacementModeMapping || source.ImportedFrom == nil {
+				t.Fatalf("helper source = %#v", source)
+			}
+			helperImports++
+		}
+	}
+	if itemImports != 3 || helperImports != 1 {
+		t.Fatalf("item imports = %d, helper imports = %d, sources = %#v", itemImports, helperImports, result.Sources())
+	}
+	if result.Config.Prompts[1].ID != "curated" {
+		t.Fatalf("helper did not provide prompt ID: %#v", result.Config.Prompts[1])
+	}
+}
+
+func TestPromptCatalogMountsResourceFromScheduledPromptOrigin(t *testing.T) {
+	result, err := consumers.LoadPromptCatalog(fixture("prompt-catalog", "config.yml"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if result.Config.Prompts[2].ID != "welcome" || result.Config.Prompts[2].Resource != "welcome.txt" {
+		t.Fatalf("scheduled resource prompt = %#v", result.Config.Prompts)
+	}
+	resourcePath := configkit.Keys("prompts").Index(2).Key("resource")
+	info, ok := result.Metadata().Lookup(resourcePath)
+	if !ok || len(info.Origin.Locations) != 1 {
+		t.Fatalf("resource origin = %#v, %v", info, ok)
+	}
+	context, ok := result.SourceContext(info.Origin.Locations[0].SourceID)
+	if !ok || context.FS == nil || filepath.Base(context.BaseDir) != "prompts" {
+		t.Fatalf("resource source context = %#v, %v", context, ok)
+	}
+	contents, err := fs.ReadFile(context.FS, pathpkg.Join(context.BaseDir, result.Config.Prompts[2].Resource))
+	if err != nil || string(contents) != "Welcome resource text.\n" {
+		t.Fatalf("mounted resource = %q, %v", contents, err)
+	}
+}
+
+func TestRequiredPromptCatalogUsesOneRequiredFile(t *testing.T) {
+	result, err := consumers.LoadRequiredPromptCatalog(fixture("prompt-catalog", "simple.yml"))
+	if err != nil || len(result.Config.Prompts) != 1 || result.Config.Prompts[0].ID != "required" {
+		t.Fatalf("simple required prompt = %#v, %v", result, err)
+	}
+}
+
+func TestPromptCatalogLeavesDomainValidationWithConsumer(t *testing.T) {
+	_, err := consumers.LoadPromptCatalog(fixture("prompt-catalog", "invalid.yml"))
+	var loadError *configkit.LoadError
+	if !errors.As(err, &loadError) || len(loadError.Diagnostics()) != 1 ||
+		loadError.Diagnostics()[0].Code != configkit.CodeValidateError ||
+		loadError.Diagnostics()[0].Path.String() != "/prompts/0/messages" {
+		t.Fatalf("consumer domain validation error = %v", err)
+	}
+}

examples/consumers/source_version.go

@@ -0,0 +1,64 @@
+package consumers
+
+import (
+	"errors"
+	"io/fs"
+
+	configkit "gitea.maximumdirect.net/eric/ConfigKit"
+)
+
+// VersionedServiceConfig demonstrates a mounted main input whose source-local
+// schema version is retained at the effective service path.
+type VersionedServiceConfig struct {
+	Service struct {
+		Version  string `yaml:"version"`
+		Name     string `yaml:"name"`
+		Endpoint string `yaml:"endpoint"`
+		Metadata struct {
+			Release string `yaml:"release"`
+		} `yaml:"metadata"`
+	} `yaml:"service"`
+}
+
+// LoadVersionedService checks the main and underlay declarations with one
+// application policy. The local rule path is "version" for each source; the
+// main input's destination places its protected value at /service/version.
+// Optional CLI overrides are passed through to ConfigKit for normal protection.
+func LoadVersionedService(mode configkit.SourceVersionMode, overrides ...configkit.Override) (*configkit.Result[VersionedServiceConfig], error) {
+	files, err := fs.Sub(bundledFiles, "testdata/source-version")
+	if err != nil {
+		return nil, err
+	}
+	main, err := fs.ReadFile(files, "main.yml")
+	if err != nil {
+		return nil, err
+	}
+	underlay, err := fs.ReadFile(files, "underlay.yml")
+	if err != nil {
+		return nil, err
+	}
+	definition := configkit.Definition[VersionedServiceConfig]{
+		Version: &configkit.VersionRule{
+			Path: configkit.Keys("version"),
+			Check: func(version configkit.Version, _ *configkit.Validation) error {
+				if !version.Present || version.Null || version.Text == "" {
+					return errors.New("a nonempty source version is required")
+				}
+				return nil
+			},
+		},
+		Secret: func(path configkit.Path) bool {
+			return path.String() == "/service/version"
+		},
+	}
+	return configkit.LoadLayered(configkit.Input{
+		Source: configkit.Bytes("service main", main), Destination: configkit.Keys("service"),
+	}, configkit.LayeredOptions[VersionedServiceConfig]{
+		Definition:        definition,
+		SourceVersionMode: mode,
+		Overrides:         overrides,
+		Underlays: []configkit.Underlay{{Inputs: []configkit.Input{{
+			Source: configkit.Bytes("service defaults", underlay), Destination: configkit.Keys("service"),
+		}}}},
+	})
+}

examples/consumers/testdata/collection-bundle/collections/weather/collection.yml

@@ -0,0 +1,9 @@
+configkit:
+  imports:
+    - path: metadata.yml
+      destination: []
+    - path: prompts/index.yml
+      destination: [prompts]
+    - path: profiles/index.yml
+      destination: [profiles]
+name: Weather Reporter

examples/consumers/testdata/collection-bundle/collections/weather/metadata.yml

@@ -0,0 +1,4 @@
+configkit:
+  imports:
+    - path: version.yml
+description: Imported collection metadata.

examples/consumers/testdata/collection-bundle/collections/weather/profiles/compact.yml

@@ -0,0 +1,3 @@
+compact:
+  format: text
+  temperature: 0.2

examples/consumers/testdata/collection-bundle/collections/weather/profiles/index.yml

@@ -0,0 +1,4 @@
+configkit:
+  imports:
+    - path: compact.yml
+    - path: weather.yml

examples/consumers/testdata/collection-bundle/collections/weather/profiles/weather.yml

@@ -0,0 +1,3 @@
+weather:
+  extends: compact
+  temperature: 0.4

examples/consumers/testdata/collection-bundle/collections/weather/prompts/brief.yml

@@ -0,0 +1,7 @@
+brief:
+  version: "1"
+  default_profile: compact
+  content_file: welcome.txt
+  messages:
+    - role: user
+      content: Write a concise weather summary.

examples/consumers/testdata/collection-bundle/collections/weather/prompts/custom.yml

@@ -0,0 +1,3 @@
+custom:
+  extends: brief
+  default_profile: weather

examples/consumers/testdata/collection-bundle/collections/weather/prompts/index.yml

@@ -0,0 +1,4 @@
+configkit:
+  imports:
+    - path: brief.yml
+    - path: custom.yml

examples/consumers/testdata/collection-bundle/collections/weather/prompts/welcome.txt

@@ -0,0 +1 @@
+Welcome to the weather collection.

examples/consumers/testdata/collection-bundle/collections/weather/version.yml

@@ -0,0 +1 @@
+version: "2.4.0"

examples/consumers/testdata/collection-bundle/operators/weather-prompt-version.yml

@@ -0,0 +1,2 @@
+brief:
+  version: "2"

examples/consumers/testdata/inheritance/presets.yml

@@ -0,0 +1,10 @@
+presets:
+  base:
+    enabled: false
+    settings:
+      primary: {width: 10}
+      secondary: {extends: primary}
+  custom:
+    extends: base
+    settings:
+      primary: {width: 20}

examples/consumers/testdata/layered-bundle/defaults/config.yml

@@ -0,0 +1,7 @@
+configkit:
+  imports:
+    - path: ../shared.yml
+enabled: true
+workers: 4
+templates:
+  welcome: messages/welcome.txt

examples/consumers/testdata/layered-bundle/defaults/messages/welcome.txt

@@ -0,0 +1 @@
+Welcome from the bundle.

examples/consumers/testdata/layered-bundle/shared.yml

@@ -0,0 +1,2 @@
+labels:
+  origin: bundle

examples/consumers/testdata/layered-main/config.yml

@@ -0,0 +1,3 @@
+mode: production
+enabled: false
+workers: 0

examples/consumers/testdata/layered-main/invalid.yml

@@ -0,0 +1 @@
+mode: unsupported

examples/consumers/testdata/native-source/config.yml

@@ -0,0 +1,2 @@
+name: embedded
+enabled: false

examples/consumers/testdata/ordinary-imports/config.yml

@@ -0,0 +1,6 @@
+configkit:
+  imports:
+    - path: metadata.yml
+    - path: prompts/brief.yml
+      destination: [prompts]
+name: Root import example

examples/consumers/testdata/ordinary-imports/metadata.yml

@@ -0,0 +1,2 @@
+version: "1.0"
+description: Metadata imported into the root document.

examples/consumers/testdata/ordinary-imports/prompts/brief.yml

@@ -0,0 +1,2 @@
+brief:
+  summary: A root-loaded prompt.

examples/consumers/testdata/prompt-catalog/config.yml

@@ -0,0 +1,15 @@
+configkit:
+  imports:
+    - dir: prompts
+      exclude: [generated-ignored.yml]
+      destination: [prompts]
+      placement: sequence-item
+    - dir: featured
+      pattern: [weather-*.yml]
+      destination: [prompts]
+      placement: sequence-item
+    - dir: optional
+prompts:
+  - id: main
+    name: Main authored prompt
+    messages: [Use the main prompt.]

examples/consumers/testdata/prompt-catalog/featured/weather-feature.yml

@@ -0,0 +1,4 @@
+id: weather
+name: Weather forecast
+messages:
+  - Describe today's forecast.

examples/consumers/testdata/prompt-catalog/invalid.yml

@@ -0,0 +1,4 @@
+prompts:
+  - id: invalid
+    name: Invalid prompt
+    messages: []

examples/consumers/testdata/prompt-catalog/optional/ignored.txt

@@ -0,0 +1 @@
+This file does not match the application import defaults.

examples/consumers/testdata/prompt-catalog/prompts/curated.yml

@@ -0,0 +1,6 @@
+configkit:
+  imports:
+    - path: helpers/curated-id.yml
+name: Curated summary
+messages:
+  - Summarize the supplied content.

examples/consumers/testdata/prompt-catalog/prompts/generated-ignored.yml

@@ -0,0 +1,3 @@
+id: ignored
+name: Excluded generated prompt
+messages: [This record is excluded.]

examples/consumers/testdata/prompt-catalog/prompts/generated-welcome.yml

@@ -0,0 +1,5 @@
+id: welcome
+name: Welcome
+messages:
+  - Welcome the user.
+resource: welcome.txt

examples/consumers/testdata/prompt-catalog/prompts/helpers/curated-id.yml

@@ -0,0 +1 @@
+id: curated

examples/consumers/testdata/prompt-catalog/prompts/required.yml

@@ -0,0 +1,3 @@
+id: required
+name: Required prompt
+messages: [Use the required prompt.]

examples/consumers/testdata/prompt-catalog/prompts/welcome.txt

@@ -0,0 +1 @@
+Welcome resource text.

examples/consumers/testdata/prompt-catalog/simple.yml

@@ -0,0 +1,7 @@
+configkit:
+  imports:
+    - path: prompts/required.yml
+      required: true
+      destination: [prompts]
+      placement: sequence-item
+prompts: []

examples/consumers/testdata/source-version/main.yml

@@ -0,0 +1,4 @@
+version: service-v2
+name: example service
+metadata:
+  release: application-2026.10

examples/consumers/testdata/source-version/underlay.yml

@@ -0,0 +1,2 @@
+version: catalog-v1
+endpoint: https://service.example.invalid