Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Evidence Guard

    2

    Fact-check AI-generated output before it ships: verifies every factual and technical claim against its real source and flags anything unsupported.

    $14

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    2 usesSecurity scanned
    Evidence Guard

    Evidence Guard

    Example session with this skill installed

    Verify the Node version and performance claims in the README and docs/api.md against our package.json and bench results before I ship this release.

    • Read your context and instructions
    • Compiled the evidence guard

    VERIFICATION NOTE — 5 claims checked.
    CRITICAL: README:12 — "Supports Node 16" — engines in package.json requires >=18.0.0.
    WARNING: docs/api.md — "ultra-low latency" — No benchmark data found to support quantitative claim.
    VERDICT: FAIL — 1 critical version mismatch.

    engines in package.json

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

    About this skill

    evidence-guard brings regulated-industry evidence standards — the same rigor used to pass medical and scientific MLR and peer review — to any AI-generated output. Before your agent ships documentation, READMEs, PR descriptions, API references, changelogs, or technical claims, evidence-guard forces it to prove every assertion against a real, verifiable source.

    The problem: AI agents write confident, plausible text that quietly contains unsupported claims — wrong version numbers, deprecated API references, performance figures with no benchmark, and documentation that drifts away from the actual code. A single model reviewing its own work rarely catches these, because the same bias that produced the claim approves it.

    evidence-guard fixes this at the level of substance, not just process. It runs a structured Claims-QC pass: it extracts every factual, technical, and quantitative claim from the output, classifies each one, verifies it against the repo or a citable source, grades the strength of the evidence, and flags risk patterns like version mismatches and doc-vs-code drift. Nothing passes the verdict gate until every critical claim is traceable. The result is a short, audit-ready Verification Note you can drop into a PR or a docs review.

    Built by PubsProToolkit, applying the evidence discipline of medical and scientific publishing to everyday agent output.

    • Cannot run code for runtime behavior.
    • Requires read access to files/sources.
    • Opinion-based claims are advisory and cannot be hard-verified.

    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

    2 uses

    Times used to date

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 5 months ago

    Requires

    Python

    Permissions

    No special permissions declared or detected

    Listed5 months ago

    What's inside

    Frequently Asked Questions

    Evidence-guard performs a "Claims-QC" pass, extracting every factual, technical, and quantitative assertion from AI output and verifying them against your codebase or citable documentation to eliminate hallucinations and version drift.

    It is designed to be model-agnostic and can be integrated into most AI agent workflows, CI/CD pipelines, or PR review bots that output technical documentation, API references, or changelogs.

    You receive the full evidence-guard skill definition, integration documentation, and a set of standardized templates for generating "Verification Notes" used in audit-ready pull requests.

    The skill identifies and classifies claims—such as version numbers, performance benchmarks, and API syntax—and flags them as "Verified," "Unsupported," or "Drift Detected" based on real source evidence.

    While powerful out of the box, you can configure the "verdict gate" threshold to decide which claim types (e.g., security-critical vs. stylistic) require the highest levels of evidence before passing.

    New in Technical Documentation