Use contextual descriptors for entity reconciliation

internal/modules/dnd/normalize/itemregistry/normalizer.go

@@ -23,7 +23,7 @@ import (
 const (
 	Key                   = "dnd/item-registry"
 	PromptID              = "dnd.item_registry.normalize"
-	normalizationPolicy   = "dnd.item_registry.normalize.v1"
+	normalizationPolicy   = "dnd.item_registry.normalize.v2"
 	semanticContextPolicy = "dnd.entity_reconcile.context.v1"
 	semanticContextRadius = 2
 	NormalizationPolicy   = normalizationPolicy

internal/modules/dnd/normalize/itemregistry/normalizer_test.go

@@ -5,6 +5,7 @@ import (
 	"encoding/json"
 	"errors"
 	"reflect"
+	"strconv"
 	"strings"
 	"testing"
 	"time"
@@ -101,7 +102,7 @@ func TestNormalizeAppliesSafeAliasProposal(t *testing.T) {
 		t.Fatalf("merged item = %#v, warnings = %#v", merged, result.Warnings)
 	}
 	encoded := string(client.requests[0].Inputs["candidates"].Content) + string(client.requests[0].Inputs["transcript"].Content)
-	if strings.Contains(encoded, doc.ID) || !strings.Contains(encoded, "candidate-000001") || strings.Contains(encoded, merged.ID) {
+	if strings.Contains(encoded, doc.ID) || strings.Contains(encoded, "candidate-") || strings.Contains(encoded, merged.ID) || !strings.Contains(encoded, `"source_refs"`) {
 		t.Fatalf("private inputs = %s", encoded)
 	}
 }
@@ -261,7 +262,7 @@ func TestRegisterPromptAssetsPreparesItemNormalizationPrompt(t *testing.T) {
 	if err != nil {
 		t.Fatal(err)
 	}
-	prepared, err := engine.Prepare(context.Background(), promptkit.RunRequest{PromptID: PromptID, PromptVersion: entityreconcile.SchemaVersion, ProfileID: "item-normalize-test", Inputs: map[string]promptkit.ArtifactRef{"candidates": promptkit.Inline(`{"candidates":[{"key":"candidate-000001","name":"Rope","source_refs":[]}]}`), "transcript": promptkit.Inline(`{"windows":[{"units":[]}]}`)}})
+	prepared, err := engine.Prepare(context.Background(), promptkit.RunRequest{PromptID: PromptID, PromptVersion: entityreconcile.SchemaVersion, ProfileID: "item-normalize-test", Inputs: map[string]promptkit.ArtifactRef{"candidates": promptkit.Inline(`{"candidates":[{"name":"Rope","source_refs":[{"start_unit_id":1,"end_unit_id":1}]}]}`), "transcript": promptkit.Inline(`{"windows":[{"units":[]}]}`)}})
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -285,10 +286,51 @@ func (c *recordingNormalizerClient) CompleteStructured(_ context.Context, reques
 	if response == "" {
 		response = `{"duplicate_groups":[]}`
 	}
-	if err := json.Unmarshal([]byte(response), output); err != nil {
+	content, err := contextualProposalResponse(response, request.Inputs["candidates"].Content)
+	if err != nil {
 		return contracts.StructuredCompletionResponse{}, err
 	}
-	return contracts.StructuredCompletionResponse{Content: json.RawMessage(response)}, nil
+	if err := json.Unmarshal(content, output); err != nil {
+		return contracts.StructuredCompletionResponse{}, err
+	}
+	return contracts.StructuredCompletionResponse{Content: content}, nil
+}
+
+func contextualProposalResponse(response string, candidateContent []byte) ([]byte, error) {
+	if !strings.Contains(response, "candidate-") {
+		return []byte(response), nil
+	}
+	var selection struct {
+		DuplicateGroups []struct {
+			Members   []string `json:"members"`
+			Canonical string   `json:"canonical"`
+		} `json:"duplicate_groups"`
+	}
+	if err := json.Unmarshal([]byte(response), &selection); err != nil {
+		return nil, err
+	}
+	var candidates struct {
+		Candidates []entityreconcile.Selector `json:"candidates"`
+	}
+	if err := json.Unmarshal(candidateContent, &candidates); err != nil {
+		return nil, err
+	}
+	selector := func(key string) entityreconcile.Selector {
+		index, err := strconv.Atoi(strings.TrimPrefix(key, "candidate-"))
+		if err != nil || index < 1 || index > len(candidates.Candidates) {
+			return entityreconcile.Selector{Name: key, SourceRefs: []entityreconcile.SourceRange{}}
+		}
+		return candidates.Candidates[index-1].Clone()
+	}
+	proposal := entityreconcile.ProposalResponse{DuplicateGroups: make([]entityreconcile.DuplicateGroup, len(selection.DuplicateGroups))}
+	for index, group := range selection.DuplicateGroups {
+		members := make([]entityreconcile.Selector, len(group.Members))
+		for memberIndex, key := range group.Members {
+			members[memberIndex] = selector(key)
+		}
+		proposal.DuplicateGroups[index] = entityreconcile.DuplicateGroup{Members: members, Canonical: selector(group.Canonical)}
+	}
+	return json.Marshal(proposal)
 }
 
 func newNormalizer(t *testing.T, client contracts.StructuredLLMClient) *Normalizer {