feat(memory): add experimental qmd search backend

docs/plans/2026-02-06-openclaw-feature-gap-analysis.md

@@ -123,7 +123,7 @@ Flynn has **6 of ~15 channels** (Telegram, WhatsApp, Discord, Slack, WebChat, TU
 | `memory.write` tool | Write memory files | Full (write/append to namespace) | **MATCH** |
 | Vector embeddings | OpenAI/Gemini/local | Full (OpenAI, Gemini, Ollama, LlamaCpp providers) | **MATCH** |
 | Hybrid search (BM25 + vector) | Full | Full (keyword + vector with configurable hybrid weight) | **MATCH** |
-| QMD backend | Experimental | -- | **MISSING** |
+| QMD backend | Experimental | Full (experimental markdown-native backend configurable via `memory.qmd`) | **MATCH** |
 
 ---
 
@@ -313,7 +313,6 @@ All five Tier 3 items implemented: Lane Queue (per-session FIFO in gateway), cre
 - Elevated mode — sandbox escape hatch
 - ~~Onboard wizard — guided setup~~ (DONE — `flynn setup` + first-run auto-trigger, 2026-02-10)
 - ClawHub/skill registry — community marketplace
-- QMD backend — experimental memory search
 
 ---
 

docs/plans/2026-02-15-openclaw-gap-roadmap.md

@@ -39,9 +39,9 @@ A gap item is considered implemented when:
 
 - Canvas / A2UI (agent-driven visual workspace)
 
-### Memory (MISSING)
+### Memory
 
-- QMD backend (experimental)
+- QMD backend (experimental) — completed on 2026-02-16
 
 ### Security (MISSING)
 

docs/plans/2026-02-16-qmd-backend-checklist.md

@@ -0,0 +1,39 @@
+# QMD Backend Checklist
+
+Date: 2026-02-16
+Status: completed
+
+## Scope
+
+- Add an experimental QMD (query markdown database) backend for `memory.search`.
+- Enable config-driven backend selection between hybrid embeddings, QMD, and keyword fallback.
+- Update docs and plan state.
+
+## Completed
+
+- Added `memory.qmd` config schema in `src/config/schema.ts`:
+  - `enabled` (default `false`)
+  - `top_k` (default `8`)
+  - `min_score` (default `0.15`)
+- Implemented `QmdSearch` backend in `src/memory/qmd-search.ts`:
+  - heading-aware scoring
+  - token overlap + phrase bonus ranking
+  - session namespace recency boost
+- Wired backend selection in `src/daemon/memory.ts`:
+  - embedding enabled -> hybrid backend
+  - else if qmd enabled -> QMD backend
+  - else keyword-only search
+- Generalized memory search tool wiring:
+  - introduced shared backend interface for `memory.search`
+  - updated memory tool factory to accept any backend implementing `search(query, topK?)`
+- Updated docs:
+  - README memory section now documents QMD config and backend precedence.
+  - OpenClaw gap docs updated to mark QMD backend as implemented.
+- Added tests:
+  - `src/memory/qmd-search.test.ts`
+  - `src/config/schema.test.ts` coverage for `memory.qmd`
+
+## Verification
+
+- `pnpm test:run src/config/schema.test.ts src/memory/qmd-search.test.ts`
+- `pnpm typecheck`

docs/plans/state.json

@@ -223,6 +223,31 @@
       ],
       "test_status": "pnpm test:run src/channels/registry.test.ts src/gateway/handlers/handlers.test.ts + pnpm typecheck passing"
     },
+    "qmd-backend": {
+      "file": "2026-02-16-qmd-backend-checklist.md",
+      "status": "completed",
+      "date": "2026-02-16",
+      "updated": "2026-02-16",
+      "summary": "Added an experimental markdown-native QMD backend for memory.search with config-driven backend selection (hybrid embeddings -> QMD -> keyword fallback), tests, and docs updates.",
+      "files_created": [
+        "docs/plans/2026-02-16-qmd-backend-checklist.md",
+        "src/memory/qmd-search.ts",
+        "src/memory/qmd-search.test.ts"
+      ],
+      "files_modified": [
+        "src/config/schema.ts",
+        "src/config/schema.test.ts",
+        "src/daemon/memory.ts",
+        "src/memory/hybrid-search.ts",
+        "src/memory/index.ts",
+        "src/tools/builtin/index.ts",
+        "src/tools/builtin/memory-search.ts",
+        "README.md",
+        "docs/plans/2026-02-06-openclaw-feature-gap-analysis.md",
+        "docs/plans/2026-02-15-openclaw-gap-roadmap.md"
+      ],
+      "test_status": "pnpm test:run src/config/schema.test.ts src/memory/qmd-search.test.ts + pnpm typecheck passing"
+    },
     "skill-safety-scanner": {
       "file": "2026-02-15-skill-safety-scanner-checklist.md",
       "status": "completed",
@@ -2267,12 +2292,12 @@
     "tier2_completion": "4/4 (100%) — inbound webhooks, vector memory search, Dockerfile, heartbeat monitor",
     "tier3_completion": "5/5 (100%) — lane queue, credential redaction, web UI token dashboard, xAI (Grok) provider, Voyage AI embeddings",
     "tier4_completion": "4/4 (100%) — gateway lock, shell completion, Tailscale Serve/Funnel, DM pairing codes",
-    "feature_gap_scorecard": "107/128 match (84%), 0 partial (0%), 21 missing (16%)",
+    "feature_gap_scorecard": "108/128 match (84%), 0 partial (0%), 20 missing (16%)",
     "operator_dx_milestone": "Phase 3 (Live Ops Dashboard): 2/2 plans complete — milestone done",
     "gmail_auth_cli": "flynn gmail-auth command implemented with OAuth2 flow, doctor check, config routed to Telegram",
     "native_audio_support": "completed — smart routing for native audio (Gemini/OpenAI/GitHub) vs Whisper transcription fallback",
     "remaining_phases_completion": "Phase 1: 3/3 (100%) — context levels, command registry, memory structure. Phase 2: 3/3 (100%) — component registry, confidence routing, history index. Phase 3: 2/2 (100%) — adaptive memory/compaction, truthfulness/autonomy hardening",
-    "next_up": "Pick the next OpenClaw gap milestone and create a scoped checklist (candidates: QMD backend, ClawHub registry, Bonjour/mDNS discovery)"
+    "next_up": "Pick the next OpenClaw gap milestone and create a scoped checklist (candidates: ClawHub registry, Bonjour/mDNS discovery, synthetic provider)"
   },
   "soul_md_and_cron_create": {
     "date": "2026-02-11",