Write workspace checkpoints during runs

docs/config.md

@@ -352,8 +352,11 @@ casts still must be present in the source transcript.
 - `resume.enabled`: boolean resume checkpointing setting. Default: `false`.
 - `debug.enabled`: boolean debug artifact setting. Default: `false`.
 
-The current run workflow uses workspace settings for diagnostics configuration.
-Checkpoint and debug artifact writers are not part of the current workflow.
+When `workspace.resume.enabled` is true, runs write stage-owned checkpoint
+artifacts under `<workspace.directory>/checkpoints/`. Checkpoint reads and
+resume execution are not implemented.
+
+Debug artifact writers are not part of the current workflow.
 
 ## Diagnostics
 

docs/internal/overview.md

@@ -42,9 +42,11 @@ production modules.
 - `internal/framework/contracts`: interfaces and request/result structs for
   input adapters, chunkers, extractors, mergers, normalizers, validators, output
   encoders, and structured LLM clients.
+- `internal/framework/checkpoint`: workspace-backed checkpoint recorder and
+  checkpoint payload envelope serialization.
 - `internal/framework/pipeline`: module registries, module specs, profile
-  resolution, capability checks, run orchestration, warnings, validation, and
-  manifest population.
+  resolution, capability checks, run orchestration, checkpoint recorder
+  boundaries, warnings, validation, and manifest population.
 - `internal/framework/llm`: Scriptorium-backed structured-output client,
   prompt/schema asset registry, scheduler, schema registry, and secret
   redaction.

docs/internal/pipeline.md

@@ -71,6 +71,12 @@ to chunk, extract, merge, and normalize requests; LLM-backed modules forward it
 through their structured completion requests so Scriptorium can include it in
 prompt execution metadata.
 
+When workspace resume checkpointing is enabled, the CLI constructs a checkpoint
+recorder after pipeline resolution and reference materialization and passes it
+through `pipeline.RunInput`. The runner records source, chunk, extract, merge,
+and normalize outcomes through that interface. Concrete modules do not receive
+workspace paths and do not write checkpoint files directly.
+
 ## Registries And Module Specs
 
 `pipeline.Registries` holds concrete constructors for execution. A

docs/operations.md

@@ -84,6 +84,24 @@ Implemented diagnostics artifacts:
 `source-document.json` is supported by the diagnostics writer but is not written
 by the current CLI run workflow.
 
+## Checkpoints
+
+When `workspace.resume.enabled: true` and `workspace.directory` is set, runs
+write checkpoints under:
+
+```text
+<workspace.directory>/checkpoints/<pipeline-id>/<input-key>-<source-digest>/<pipeline-digest>/
+```
+
+Each workflow step owns its own manifest and payload files. There is no
+root-level checkpoint summary. Current runs write checkpoints for inspection and
+future recovery support only; the CLI does not read checkpoints or skip work.
+
+Checkpoint payloads preserve byte content with base64 envelopes, media type,
+metadata, warnings, and content digests where applicable. Checkpoints do not
+include raw prompts, raw reference contents, raw LLM request payloads, or debug
+traces.
+
 ## Retention
 
 Diagnostics retention is configured with `workspace.diagnostics.retention`,
@@ -138,8 +156,8 @@ directories unless they are part of your own operational policy.
 
 ## Operational Limits
 
-There is no command to resume a failed run. Re-run `notarius run` after fixing
-the cause.
+Checkpoint writing does not provide resume execution yet. Re-run `notarius run`
+after fixing the cause of a failed run.
 
 Provider retries and timeouts are handled by Scriptorium according to the
 selected execution profile. Pipeline module retries are controlled by module