Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    MCP FactCheck Operator

    1

    Use when checking whether a statement is true, auditing a draft for unsupported claims, or attaching published ratings to a document.

    $5.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    MCP FactCheck Operator

    MCP FactCheck Operator

    Example session with this skill installed

    Check this landing page claim before we publish: studies show remote workers are 13% more productive, which is why 40% of US companies have gone fully remote.

    • Read your context and instructions
    • Compiled the mcp factcheck operator

    Two claims split before any call. The productivity figure returns a published review rated Missing context, quoted with publisher and date and reported as incomplete rather than false. The 40% figure returns no review, and the fetched page does not carry it. Reported as one claim qualified by a review and one unsupported as worded, with the recommended edit, the coverage gap, and a call log.

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

    About this skill

    Drives all six tools on the factcheck MCP server — check_facts_google, verify_claim, cross_verify, fetch_source, get_citations, anti_hallucination_protocol — to verify claims before publishing, submitting, or shipping. Use when checking whether a statement is true, auditing a draft for unsupported claims, or attaching published ratings to a document. Teaches which instrument supports which verdict, how to read overlap scores without overclaiming, and how to report a tool that was never configured.

    Coverage is uneven: published fact checks cluster around politics, health and viral claims, so technical and local claims usually return nothing. It does not adjudicate truth. A tool that was never configured is reported as not run, never as a negative finding.

    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 19 days ago

    Requires

    Factcheck MCP
    Google Fact Check API key

    Permissions

    No special permissions declared or detected

    Listed19 days ago

    What's inside

    Frequently Asked Questions

    Five of the six tools still work: open-web corroboration, evidence scoring, page fetching, paper search and the checklist. Only the published-review lookup is skipped, and the report says so.

    No. It means no publisher reviewed that claim. The skill escalates to primary sources and leaves the verdict open.

    Because they answer six different questions. The package maps which verdict each one is allowed to produce, which is what stops an agent calling a search result verified.

    New in LLM Ops