More screenshots
Works with the AI Agents you already use
NotebookLM: Interrogate Your Sources Like an Investigator
Turn a pile of sources into defensible answers with NotebookLM — curated corpus, a question ladder from facts to critique, cross-examination, and provenance on every claim.
$5
NotebookLM: Interrogate Your Sources Like an Investigator
Example session with this skill installed
We have 12 docs about our product spec and they contradict each other. Use NotebookLM to map every requirement, surface the contradictions, and find what the corpus never addresses.
- Read your context and instructions
- Compiled the notebooklm
Curated the corpus first: 12 docs — PRD, 4 design docs, 6 meeting notes, 1 competitor teardown — tagged by date and author team, with stance noted on each. Two maybe-useful docs failed the intake checklist and stayed out; three copies of the same press release were deduped to one. The strongest counter-source was deliberately added — a corpus where everything agrees is a filter bubble. Rung 1 extraction: every stated requirement with its source doc — 34 requirements. Rung 2 comparison: 6 direct contradictions surfaced, including the PRD saying 'real-time sync' while design doc 3 says 'hourly batch acceptable for v1'. Rung 3 synthesis: 28 requirements have consensus; 6 are contested, each with the disagreeing parties named. Rung 4 critique: no source addresses migration from the legacy system — the corpus has a hole shaped exactly like the riskiest work. Cross-examination asked the same question per source and compared in a table, demanded quotes for contested claims, and probed the negative space — which sources do not mention migration, and is that surprising? Deliverable: a requirements table with per-claim sources, a contradictions section with quotes, and a gaps section. The team resolved 5 of 6 contradictions in one meeting — because the disagreement was finally visible.
Connects securely to your tools. The creator never sees your data.
About this skill
NotebookLM is only as good as the corpus you feed it and the questions you ask it — most people dump 50 PDFs and ask 'summarize', which is how confident garbage gets generated. This skill is the discipline that turns a source collection into an evidence base: curate the inputs with a stance and recency note on each, climb a question ladder from extraction through comparison and synthesis to critique before any decision support, cross-examine the same question per source, and export every claim with its provenance attached.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.
Answers are only as good as the corpus: a 15-source corpus that passes the intake checklist outperforms a 60-source dump every time.
It cannot answer what the sources do not cover — that is what the gaps section is for, and a gaps section is not a failure.
Model inference must be labeled as inference: 'Source A says X' and 'combining A and B suggests Y' are different kinds of claims and need different scrutiny.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 3 days ago
Permissions
No special permissions declared or detected