Menu
Notebook First retrieval policy — query NotebookLM notebooks before web search or context dumping
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
Apply this skill for Unite-Hub Supabase migrations, PostgREST/Data API visibility, founder-scoped Playwright journeys, or errors such as PGRST205, access=denied, stale Supabase linked refs, or migration history drift. Prevents repeating the SQL/cache/auth loop by enforcing the exact verification sequence for core journeys.
The compass for Unite-Hub's road to /shipit. Defines the single NorthStar (a real, comprehensive, working founder CRM in production, every section GREEN), the binding definition of GREEN, and the No-Invaders Manifest that keeps the build honest and surgical. Consult BEFORE deciding what to build/skip/finish — it resolves "200 ≠ real" temptations and scope-creep pressure. P1, auto-loaded.
Apply this skill WHEN scaffolding a new cron "pull" route that syncs external/derived data into Supabase on a schedule (Vercel cron). Encodes the Unite-Hub cron invariants: CRON_SECRET auth, FOUNDER_USER_ID actor, overlap safety, idempotent upsert, last-sync timestamp, and failure surfacing. Generic `cron-scheduler` covers scheduling; this covers the PULL handler body. P3.
Apply this skill WHEN verifying that a route, page, or integration serves REAL data and not silent mock/placeholder data. Detects the "false-green" failure mode: an endpoint returns 200 (or a page renders) while the underlying data is fabricated because a provider is unconnected. Trigger WHENEVER classifying a section's readiness, reviewing integration wrappers, or before marking anything GREEN. P2 — load on audit/verify tasks.
Manifest-first context isolation — each subagent receives only its scope, never the full codebase
Apply this skill for ANY decision with non-obvious tradeoffs: architectural choices, debugging without a clear root cause, performance strategies, security decisions, feature design with competing constraints, refactoring scope decisions. Forces multi-perspective analysis before committing to a solution. P1 auto-load — always active on complex reasoning tasks.
| id | notebooklm-second-brain |
| name | notebooklm-second-brain |
| type | skill |
| version | 1.0.0 |
| created | 20/03/2026 |
| modified | 20/03/2026 |
| status | active |
| category | meta-orchestration |
| tags | ["retrieval","knowledge-management","notebooklm","second-brain"] |
| triggers | ["onboarding","debugging","architecture","security","post-build"] |
| description | Notebook First retrieval policy — query NotebookLM notebooks before web search or context dumping |
Policy: Query the relevant NotebookLM notebook before web search or dumping full docs into context.
This skill enforces a "Notebook First" retrieval policy. Instead of loading entire markdown files or running web searches for architecture decisions, debugging patterns, or security knowledge, the agent queries purpose-built NotebookLM notebooks via the nlm CLI. This keeps the agent context window lean and retrieval fast.
| Scenario | Notebook | Why |
|---|---|---|
| "How does auth work?" | project_sot | Architecture decisions live here |
| "Why is this test failing?" | debug_kb | Error patterns and debugging playbooks |
| "Is this input sanitised?" | security_handbook | OWASP patterns and security reviews |
| "Explain the codebase" | repo_onboarding | Codebase atlas and mind maps |
| After successful build/verify | project_sot | Auto-sync implementation notes |
All notebook IDs are stored in .claude/notebooklm/notebooks.json. Never hard-code or fabricate IDs.
# Read current notebook config
cat .claude/notebooklm/notebooks.json
| Query Type | Target Notebook | Fallback |
|---|---|---|
| Architecture, feature specs, implementation history | project_sot | Read CLAUDE.md directly |
| Error patterns, test failures, debugging steps | debug_kb | Search codebase with grep |
| OWASP, auth patterns, security reviews | security_handbook | Web search security docs |
| "How does X work?", codebase overview, onboarding | repo_onboarding | Read source files directly |
# Authentication
nlm login # Browser cookie extraction (one-time)
nlm login --check # Verify auth status
# Notebook management
nlm notebook list # List all notebooks
nlm notebook create "Name" # Create new notebook
nlm notebook query <notebook_id> # Interactive query
# Source management
nlm source add <notebook_id> --file <path> # Add file source
nlm source add <notebook_id> --url <url> # Add URL source
nlm source add <notebook_id> --text "content" # Add text source
nlm source list <notebook_id> # List sources
# Notes and audio
nlm note create <notebook_id> "Title" # Create note
nlm audio create <notebook_id> --confirm # Generate audio overview
When setting up a new notebook:
nlm notebook create "Name"notebooks.jsonAfter successful pnpm run verify or pnpm build, the sync hook (.claude/hooks/scripts/notebooklm-sync.ps1) automatically creates an implementation note in project_sot with:
notebooks.json (populated by /notebooklm-bootstrap)| System | Integration |
|---|---|
| Beads | Sync completed Beads tasks as implementation notes |
| Council of Logic | Architecture queries route through project_sot first |
| Hooks | Post-verify sync hook adds notes automatically |
| Skills | Other skills can query notebooks for domain knowledge |
Run the bootstrap command to install, authenticate, and create all notebooks:
/notebooklm-bootstrap
This installs notebooklm-mcp-cli, authenticates via browser, creates the 4 notebooks, seeds them with project sources, and writes IDs to notebooks.json.
python scripts/validate-notebooks.py # Schema validation
python scripts/validate-notebooks.py --check-ids # Verify IDs populated
python scripts/validate-notebooks.py --dry-run-sync # Preview sync note