chore: add extraction-service to AGENTS.md, CI matrix, token pre-commit hook (7.9)

- AGENTS.md: extraction-service in layout, file ownership, deps, test table (211+ tests)
- ci.yml.disabled: extraction package + extraction-service added to matrix
- package.json: lint-staged rule to auto-generate tokens on bytelyst.tokens.json change
- ROADMAP.md: 252/278 (~91%), 7.9 done

.github/workflows/ci.yml.disabled

@@ -76,6 +76,7 @@ jobs:
           - fastify-core
           - design-tokens
           - testing
+          - extraction
 
     steps:
       - name: Checkout
@@ -124,6 +125,7 @@ jobs:
           - billing-service
           - growth-service
           - tracker-service
+          - extraction-service
 
     steps:
       - name: Checkout

AGENTS.md

@@ -42,6 +42,7 @@ learning_ai_common_plat/
 │   ├── billing-service/            # Subscriptions, Stripe, usage, licenses (port 4002)
 │   ├── growth-service/             # Invitations, referrals, promos (port 4001)
 │   ├── tracker-service/            # Items, comments, votes, public roadmap (port 4004)
+│   ├── extraction-service/         # LangExtract text extraction + Python sidecar (port 4005)
 │   └── monitoring/                 # Loki + Grafana config, health-check script
 ├── docs/                           # Architecture docs, roadmap, analysis
 ├── package.json                    # Root scripts: build, test, typecheck, clean
@@ -106,7 +107,7 @@ learning_ai_common_plat/
 ## 5. File Ownership Map
 
 | Domain                  | Location                       | Key Files                                                                                                                     |
-| ----------------------- | ---------------------------- | ----------------------------------------------------------------------------------------------------------------------------- |
+| ----------------------- | ------------------------------ | ----------------------------------------------------------------------------------------------------------------------------- |
 | **Errors**              | `packages/errors/`             | `src/index.ts` — `BadRequestError`, `UnauthorizedError`, `ForbiddenError`, `NotFoundError`, `ConflictError`, `RateLimitError` |
 | **Cosmos client**       | `packages/cosmos/`             | `src/index.ts` — `getCosmosClient()`, `getContainer()`, container registry                                                    |
 | **Config / Product ID** | `packages/config/`             | `src/index.ts` — `loadEnvConfig()`, `loadProductIdentity()`, `getProductId()`                                                 |
@@ -132,6 +133,10 @@ learning_ai_common_plat/
 | **Public roadmap**      | `services/tracker-service/`    | `src/modules/public/`                                                                                                         |
 | **Tracker comments**    | `services/tracker-service/`    | `src/modules/comments/`                                                                                                       |
 | **Tracker votes**       | `services/tracker-service/`    | `src/modules/votes/`                                                                                                          |
+| **Extraction routes**   | `services/extraction-service/` | `src/modules/extract/` — POST /extract, /extract/batch, /extract/jobs, /extract/models                                        |
+| **Extraction tasks**    | `services/extraction-service/` | `src/modules/tasks/` — predefined task library (triage, transcript, memory-insight, etc.)                                     |
+| **Extraction Python**   | `services/extraction-service/` | `python/src/` — LangExtract sidecar (FastAPI :4006), extractor, task registry, language detection                             |
+| **Extraction package**  | `packages/extraction/`         | `src/index.ts` — `createExtractionClient()`, shared types for consumers                                                       |
 | **Monitoring**          | `services/monitoring/`         | `health-check.ts`, `loki/`, `grafana/`                                                                                        |
 
 ### Dashboard Consumers (via `file:` refs)
@@ -271,6 +276,7 @@ DEFAULT_PRODUCT_ID=lysnrai
 @lysnrai/billing-service  ← @bytelyst/{config, cosmos, errors}
 @lysnrai/growth-service   ← @bytelyst/{config, cosmos, errors}
 @lysnrai/tracker-service     ← @bytelyst/{config, cosmos, errors}
+@lysnrai/extraction-service  ← @bytelyst/{fastify-core, config, cosmos, errors, auth}
 ```
 
 Build order: packages first (they have no inter-deps), then services. `pnpm build` handles this automatically via workspace topology.
@@ -289,12 +295,13 @@ Build order: packages first (they have no inter-deps), then services. `pnpm buil
 ## 11. Service Test Counts
 
 | Service / Package  | Tests    | Runner |
-| ----------------- | -------- | ------ |
+| ------------------ | -------- | ------ |
 | platform-service   | 55       | vitest |
 | tracker-service    | 45       | vitest |
+| extraction-service | 46       | vitest |
 | growth-service     | 33       | vitest |
 | billing-service    | 32       | vitest |
-| **Total**         | **165+** |        |
+| **Total**          | **211+** |        |
 
 ## 12. Common Pitfalls
 

docs/ROADMAP.md

@@ -537,7 +537,7 @@ The following gaps were identified by scanning every import in the actual codeba
 - [x] **7.6** Add `@bytelyst/testing` package (shared test utilities, mock factories) — created with 10 tests
 - [ ] **7.7** Evaluate Python shared package for `cosmos_client.py` + `blob_client.py` if MindLyst adds Python backend
 - [ ] **7.8** Integrate `@bytelyst/design-tokens` into LysnrAI dashboards (unified design language)
-- [ ] **7.9** Add pre-commit hooks to auto-run token generation when JSON changes
+- [x] **7.9** Add pre-commit hooks to auto-run token generation when JSON changes — lint-staged rule on `bytelyst.tokens.json`
 - [ ] **7.10** Set up Renovate/Dependabot for common-plat dependency updates
 
 ---
@@ -557,8 +557,8 @@ The following gaps were identified by scanning every import in the actual codeba
 | **4**     | `@bytelyst/design-tokens` (4 platforms)          | 24      | 23      | ✅ CSS synced to MindLyst; CONTRIBUTING updated; visual verify pending            |
 | **5**     | CI/CD + Docker (pre-copy strategy)               | 23      | 21      | ⚠️ All Dockerfiles rewritten, CI workflows created; Docker build blocked by proxy |
 | **6**     | Verification + docs + cleanup                    | 28      | 21      | ✅ Docs updated, depcheck done, git clean; E2E needs services                     |
-| **7**     | Future enhancements (+testing pkg)               | 10      | 3       | 🔲 @bytelyst/testing + @bytelyst/blob + @bytelyst/monitoring created              |
+| **7**     | Future enhancements (+testing pkg)               | 10      | 3       | 🔲 @bytelyst/testing (10 tests) + token pre-commit hook + AGENTS updated          |
-| **Total** | **10 packages (+1 bonus: logger)**               | **278** | **251** | **~90% complete**                                                                 |
+| **Total** | **10 packages (+1 bonus: logger)**               | **278** | **252** | **~91% complete**                                                                 |
 
 ### Bonus Package (not in original roadmap)
 

package.json

@@ -53,6 +53,9 @@
     ],
     "*.{js,jsx,json,md,yml,yaml}": [
       "prettier --write"
+    ],
+    "packages/design-tokens/tokens/bytelyst.tokens.json": [
+      "pnpm --filter @bytelyst/design-tokens generate"
     ]
   }
 }