Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Evidence Verification Pro — Check Whether Claims Are

    1

    Audit exact claim-evidence relationships for support, scope, dates, provenance, source independence, quantitative accuracy and remaining uncertainty.

    $9.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Evidence Verification Pro — Check Whether Claims Are

    Evidence Verification Pro — Check Whether Claims Are

    Example session with this skill installed

    Illustrative scenario: Claim: “Our new onboarding program doubled retention and caused the improvement.” Supplied internal note: 10 of 20 participants remained after six months this year; 5 of 20 remained in a prior cohort. No comparison method or control group is provided. Verify the claim from this evidence only.

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

    Atomic claim 1: Retention doubled between the supplied cohorts.
    Calculation: Current 10/20 = 50%; prior 5/20 = 25%. Relative ratio = 2; difference = 25 percentage points.
    Verdict: Numerically supported for the supplied counts, conditional on comparable cohort definitions and measurement periods. Independent data verification has not occurred.

    Atomic claim 2: The onboarding program caused the improvement.
    Verdict: Insufficient evidence. Two cohort counts alone do not rule out selection, staffing, timing or other changes. No control or causal method supplied.

    Source/provenance: Supplied internal note only. Source date, underlying records and approval status are unknown. It is not independent confirmation.

    Corrected wording: “In the supplied cohorts, six-month retention was 50% versus 25% previously, a 25-percentage-point increase. The data do not establish that onboarding caused the difference.”

    Missing evidence: Consistent population and retention definitions, underlying cohort records, relevant contextual changes, uncertainty assessment and a defensible evaluation design.

    Reverification path: Validate counts and comparable scope, then assess causal evidence using an appropriate method and qualified review.

    Limits: No external sources searched, private records accessed, causal effect estimated or consensus claimed.

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

    About this skill

    Evidence Verification Pro helps teams evaluate whether sources actually establish a claim. It decomposes compound assertions, identifies the required evidence, traces primary versus derivative sources, checks supporting passages and compares scope, dates, population and geography.

    The skill detects citation mismatch, circular sourcing, causal overclaims, denominator errors and conflicts. Claim-evidence ledgers state a reasoned verdict, confidence, missing evidence and a reverification path. A citation’s presence does not prove support, and repeated reporting is not independent confirmation.

    Works in ChatGPT and Claude with supplied material or actual authorized research tools. It never invents sources, access or consensus. Missing support is not automatically proof of falsehood; judgments stay limited to available evidence and qualified review for consequential domains.

    Verdicts are limited by source access, freshness and quality. Lack of evidence is not proof of falsehood. Cannot invent missing sources or certify legal, medical or financial conclusions. Current verification depends on available authorized tools.

    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 evidence-source reading and agreed verification-report writing. No external access or disclosure implied. Preserve source support and minimize confidential data; specialist review governs high-stakes conclusions.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    No. It checks whether the cited passage actually entails the claim at the correct scope and date.

    Not when they repeat the same underlying source; provenance and independence are checked.

    It returns a qualified verdict and specifies the missing evidence and reverification path rather than inventing support.

    New in Research & Analysis