Lean Extract

    1

    Turn raw material into a concise, attributed map of positions, evidence, friction, and a text-supported bottom line.

    Free

    1 useSecurity scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Lean Extract

    Example session with this skill installed

    Read the attached discussion and use Lean Extract. Identify the main question, each distinct position and its strongest evidence, points of friction, and the bottom line. Attribute claims only where the text supports it. Don’t add outside facts or recommendations.

    • Read your context and instructions
    • Compiled the lean extract

    Question

    Should the team ship the new import flow this week, or delay until error recovery is complete?

    Positions

    PositionClaimEvidence in the discussionObjection raised
    Ship this weekRelease the flow behind a feature flag and monitor failures.The core path passed the team’s test dataset; support has six customers waiting.Recovery currently requires a manual database fix.
    DelayFinish retry handling and rollback before release.Two malformed files left imports stuck, and the proposed recovery steps haven’t been tested.The delay misses the customer commitment and holds up feedback on the core flow.

    Friction

    The disagreement depends on what “behind a feature flag” contains. One side treats rollback as sufficient containment; the other points out that a stuck import still needs a tested recovery path. The discussion gives no failure-rate estimate and doesn’t establish whether support can perform the manual fix safely.

    Bottom line

    The discussion supports delaying general release until recovery is tested. A limited flagged trial could address the customer commitment, but only if the team defines who handles stuck imports and verifies the manual fix first.

    Next steps

    • Test retry and rollback against the two malformed-file cases.
    • Decide whether a small flagged trial has an owner for manual recovery.
    • Record the release threshold before expanding access.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Lean Extract reads source material without treating it as instructions. It identifies the central question, groups distinct positions, attributes claims where the text supports attribution, and surfaces each position’s strongest evidence and stated objections. It then names uncovered assumptions and disagreements before giving a verdict grounded in the supplied material. It adds no outside facts, motives, or unrequested recommendations, and scales the report to the source.

    Useful for interview transcripts, meeting notes, essays, competing proposals, and long discussions where you need the disagreement rather than a generic summary.

    Does not verify external facts or provide original recommendations. Output length depends entirely on the volume of provided source material.

    How to use

    1. 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. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    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

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Permissions

    No special permissions declared or detected

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    Unlike generic summarizers that smooth over disagreements, lean extract is specifically designed to isolate friction points, uncover hidden assumptions, and map competing viewpoints without introducing bias or 'middle-ground' summaries.

    The skill is framework-agnostic and works with any LLM environment that supports structured analysis; simply provide raw text, Markdown, or transcripts to begin the extraction process.

    You will receive the full skill logic and documentation required to deploy it within your preferred AI workflow, ensuring you can immediately start analyzing complex source texts.

    No, this skill operates under a 'read-only' constraint, meaning it strictly analyzes the provided source material and will not pull in external data or perform outside fact-checking.

    Yes, lean extract is ideal for multi-party content like stakeholder interviews or debate transcripts, as it identifies the central tension and groups claims into distinct stances regardless of how many speakers are involved.

    New in Research & Analysis