Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add MarcosNahuel/antigravity-plugin-cc --skill notebook-kbgit clone --depth 1 https://github.com/MarcosNahuel/antigravity-plugin-ccWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/marcosnahuel/antigravity-plugin-cc/notebook-kb)<a href="https://agentmods.dev/skills/marcosnahuel/antigravity-plugin-cc/notebook-kb"><img src="https://agentmods.dev/badge/skills/marcosnahuel/antigravity-plugin-cc/notebook-kb/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/marcosnahuel/antigravity-plugin-cc/notebook-kb"><img src="https://agentmods.dev/badge/skills/marcosnahuel/antigravity-plugin-cc/notebook-kb.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00108 | $0.01323 |
| Opus 5.5 | $0.00043 | $0.00529 |
| Sonnet 5.5 | $0.00022 | $0.00265 |
| Haiku 4.5 | $0.00011 | $0.00132 |
Grade A, and why
notebook-kb scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
notebook-kb — work against the local document RAG
/agy:notebook <folder> | <objective> analyzes a folder of documents and compiles a queryable
SQLite database docs/agy/notebook/<slug>/notebook.db: documents, chunks (+FTS5 / optional vectors), entities, events, relations, citations. Every fact row carries a quote and a source
document. This skill is how you USE that DB to do real work — deterministically, with citations, and
without pulling the documents back into Claude's context.
Decision gate — when to use the DB
- Use the DB (
/agy:notebook-query) for structured / aggregate / grounding work: totals of amounts by category, "which documents mention <person/org/term>", timelines, entity rosters, exporting a table, verifying a figure against its source. SQL is exact and auditable; prose is not. - Use
/agy:notebook-askfor an open-ended prose answer grounded in the summaries. - Build/refresh first if needed: if
notebook.dbis missing → run/agy:notebook <folder> | <objective>. If it's older than the newest*.facts.json→ rebuild (Phase 1.5):python "<plugin>/scripts/notebook_db.py" "<OUTDIR>" "<objective>"(~1s, pure Python).
How to query (there is NO sqlite3 CLI — always Python, read-only)
python - "<OUTDIR>/notebook.db" "<SQL>" <<'PY'
import sqlite3, sys, json
con = sqlite3.connect("file:%s?mode=ro" % sys.argv[1], uri=True); con.row_factory = sqlite3.Row
try: print(json.dumps([dict(r) for r in con.execute(sys.argv[2])], ensure_ascii=False, indent=2, default=str))
except Exception as e: print("SQL_ERROR: %s" % e)
PY
Prefer the v_* views (they dedup by ent_key and keep citations). The schema + a recetas cookbook
live in the /agy:notebook-query command file — reuse those queries. Entity taxonomy:
persona | organizacion | monto | fecha | referencia.
Citation contract (non-negotiable for trustworthy answers)
- Every claim cites its source:
doc_ref(orbasename) of the document the row came from. - A SUM lists its contributing rows so the total is auditable line by line. Monetary math is in
integer
monto_cents; divide by 100 only to display (no float drift). - 0 rows → say "does not appear in the corpus", and surface coverage gaps:
SELECT nn,tipo,basename FROM documents WHERE estado='no_procesado'. Never invent a name, amount, date or reference — if it isn't a row in the DB, it isn't a fact.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 83 lines · 108 tokens per session scan A c4feff49dd53
notebook-kb is a skill published in the GitHub repository MarcosNahuel/antigravity-plugin-cc (31 stars, last pushed 13d ago), licensed MIT. It adds 108 tokens to every session and 1,323 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
flux-migrate
Build zero-downtime database migrations — forward SQL, rollback SQL, deployment sequence. Use when asked to "write migration", "schema change", "add column", "rename table", "drop column", or "migrate safely".
forge-infra
Build production-grade infrastructure as code for a service or project. Use when asked to "set up infra", "provision infrastructure", "create cloud resources", "IaC for this project", "terraform for this", or "deploy this service".
flux-schema
Design and build database schema — tables, columns, types, indexes, constraints, relationships. Given a domain description, output the schema and write the files. Use when asked to "design schema", "database design", "create tables", or "data model".
flux-recon
Database reconnaissance — full inventory of schema, migrations, data volume, backups, connection pooling, and query patterns. Use when asked to "assess this database", "understand the schema", or "database health check".
forge-audit
Audit existing infrastructure for security issues, waste, and misconfigurations. Use when asked to "audit my infra", "check cloud setup", "infra review", "are we wasting money", "security check on infra", or "review my terraform".
flux-health
Data quality and pipeline health check — freshness, schema drift, null rates, orphaned records, pipeline status. Use when asked about "data quality check", "pipeline health", "is our data fresh", or "schema drift".