Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Evidence Extraction Pro — Turn Papers Into Structured

    1

    Extract source-reported study details and findings into reusable evidence records with provenance, missingness and clear reported-versus-derived distinctions.

    $9.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Evidence Extraction Pro — Turn Papers Into Structured

    Evidence Extraction Pro — Turn Papers Into Structured

    Example session with this skill installed

    Illustrative source excerpt, not a real paper: “Twenty students completed a four-week writing workshop. Mean rubric score rose from 60 before the workshop to 70 afterward. No comparison group was used.” Extract a record using only this excerpt; title, authors and statistical uncertainty are absent.

    • Read your context and instructions
    • Compiled the evidence extraction pro

    Source status: Supplied fictional excerpt only; not a verified publication or full-text study.

    Study identity: Title, authors, year, DOI and registry identifier — not reported. No citation fabricated.
    Population/sample: Twenty students completed the workshop. Enrolled number and exclusions are not reported; do not equate completers with total recruitment.
    Intervention: Four-week writing workshop. Content and delivery details not reported.
    Comparator: No comparison group, explicitly reported.
    Outcome: Rubric score; rubric definition and scale range not reported.
    Timepoints: Before and after the workshop; exact measurement dates not reported.
    Reported results: Mean before = 60; mean after = 70.
    Derived value: Mean difference = 70 − 60 = +10 rubric-score units, calculated from the excerpt. This is derived, not a reported effect estimate with uncertainty.

    Uncertainty: Standard deviations, confidence intervals, statistical tests and participant-level data not reported. No standardized effect or significance can be established.
    Interpretation: Scores increased among the described completers, but the uncontrolled before/after excerpt does not establish causal workshop effectiveness.
    Provenance: Supplied excerpt sentences 1–3; document page and section unavailable.
    Extraction confidence: High for the literal supplied values; incomplete for study identity, design detail and broader conclusions.

    Next step: Obtain the authorized full source and supplements to verify identity, definitions, analysis population and missing methods. No literature search, quality rating or meta-analysis performed.

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

    About this skill

    Evidence Extraction Pro helps researchers and evidence teams convert papers, abstracts, reports, tables and supplements into structured findings. It preserves study identity, populations, methods, interventions or exposures, comparators, outcomes, timepoints, results and uncertainty without creatively filling gaps.

    The skill organizes evidence matrices and extraction records, links related reports when supported, normalizes compatible terms and units, and flags contradictions across text, tables and supplements. Consequential values retain source provenance; reported facts, author interpretation and extractor inference remain separate.

    Works in ChatGPT and Claude with authorized source documents or legitimate accessible research tools. Abstract-only work is labeled and source access limits confidence. It does not invent citations, page numbers, sample sizes, statistics or missing methods, and extraction is not proof of study quality or causality.

    Missing source fields stay unknown. OCR, tables and figures can be ambiguous; full-text access and human verification may be needed. Extraction does not establish causality or study quality. Cannot invent statistical uncertainty or source locations.

    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

    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 1 month ago

    Permissions

    Read Files
    Write Files

    File Scopes

    Authorized
    task
    inputs
    and
    agreed
    output
    directory
    only

    Optional authorized scholarly-source reading and agreed extraction-record writing. Use lawful accessible material and preserve provenance. No inaccessible database access or clinical decision-making implied. Qualified researchers review extraction and synthesis.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    No. Absent fields remain not reported or unknown; conventions are not evidence.

    Yes, with abstract-only access clearly labeled and confidence limited to what is actually reported.

    No. Derived values include their calculation and are explicitly distinguished from source-reported findings.

    New in Research & Analysis