[CLAUDE] Docs: setup RAG Framework v1.3 governance + eval framework

- docs/governance/README.md: Path B delegation stub → AI_INFRA canonical
  Phase/BC vocabulary documented (9 phase + 10 BC SOLUTION_ERP-specific)
- .claude/rag.json: add _decision_log block (10 rationale entries) +
  add .claude/agents/**/*.md to corpus_paths (fix Case D harvest gap)
- eval/evaluator.md: inline executor spec v1.0 (Spec A strict)
- eval/golden-set-solution_erp.jsonl: 14-entry golden set v1.1
  (5 gotcha + 3 pattern + 3 decision + 3 negative)
- eval/runs/2026-05-26-baseline-v1.0-failed.json: v1.0 attempt
  recall@5=0.455 FAIL — root cause diagnosis Case A/C/D
- eval/runs/2026-05-26-baseline-v1.1-pending.json: v1.1 attempt
  pending CLI restart for accurate numbers
- eval/trial-state-lock.json: 2-section split (quality_gate +
  drift_monitor) per v1.3 §6.2, 4-week milestones 2026-05-26 → 2026-06-23

CRITICAL lesson: bootstrap.py --project flag overrides collection name only.
Use --config D:\...\SOLUTION_ERP\.claude\rag.json for correct project root.
Old projects.json had root_path=AI_INFRA for solution_erp (Anti #24) — FIXED.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

.claude/rag.json

@@ -4,7 +4,8 @@
   "corpus_paths": [
     "docs/**/*.md",
     ".claude/agent-memory/**/MEMORY.md",
-    ".claude/skills/**/SKILL.md"
+    ".claude/skills/**/SKILL.md",
+    ".claude/agents/**/*.md"
   ],
   "exclude_paths": [
     "docs/_archive/**",
@@ -19,7 +20,25 @@
   "share_to_global": true,
   "search_scope_default": ["self", "shared_global"],
   "auto_reindex": true,
+  "auto_reindex_mode": "replace",
   "chunk_size": 1500,
   "chunk_overlap": 200,
-  "contextual_retrieval": true
+  "contextual_retrieval": true,
+  "_decision_log": {
+    "auto_reindex_mode_rationale": "Layer C mandate per VIPIX commit c029ddb (gotcha #9 corpus drift fix forward). replace = full re-index each session, prevents append duplicate accumulation. DO NOT deviate.",
+    "share_to_global_rationale": "true — SOLUTION_ERP patterns proven cross-project applicable (gotcha catalog, CQRS pattern, multi-agent setup). NO PII/creds in corpus (docs + skills + agent-memory only).",
+    "chunk_size_rationale": "Legacy 1500/200 retained 2026-05-26. v1.3 §12.4 recommends 512/10-20%. Empirical retest required before flip — schedule trial week 3 (2026-06-16) evaluate.",
+    "contextual_retrieval_rationale": "Flag true but per v1.3 §12.1 + §9.4: SOLUTION_ERP chunks self-contained (gotchas, patterns, decisions) → Contextual Retrieval prepend likely wasteful. Evaluate per eval recall@5 trial week 3.",
+    "spec_a_vs_b_resolution_chosen": "Spec A — Strict. Rationale: SOLUTION_ERP chunks canonical + finite scope (51 gotchas, patterns, decisions) → strict retrieval test appropriate.",
+    "spec_chosen_date": "2026-05-26",
+    "anatomy_threshold_chosen": "6/6 STRICT per v1.3 §5.2 default (corpus 11,922 chunks — mature)",
+    "registry_drift_note": "Anti #24 — projects.json registry 2830 vs Qdrant LIVE 11,922 (+321% drift). Intentional defer re-bootstrap until Phase 9 UAT stable. Document in trial-lock _baseline_note.",
+    "source_path_note": "Anti #23 — absolute Windows path D:\\Dropbox\\... in chunk payload. Fix in next re-bootstrap via bootstrap.py path normalization. Low priority.",
+    "governance_doc": "docs/governance/README.md (Path B delegation stub — AI_INFRA canonical)"
+  },
+  "_notes": {
+    "bootstrap_command": "python D:\\Dropbox\\CONG_VIEC\\AI_INFRA\\claude-rag\\bootstrap.py --project solution_erp",
+    "cross_project_ai_infra_pointer": "AI Infra centralized at D:\\Dropbox\\CONG_VIEC\\AI_INFRA\\. Cross-project tooling (RAG/MCP/Governance) — single source of truth.",
+    "layer_c_mandate": "auto_reindex_mode=replace added 2026-05-23 broadcast per VIPIX commit c029ddb — gotcha #9 RAG corpus drift fix forward"
+  }
 }