Works with the AI Agents you already use
Fact-Check & Compliance Gate
Fact-check an article, post or report before publishing: extracts each claim, verifies it, and flags legal, regulatory and compliance risks.
$9.99
Fact-Check & Compliance Gate
Example session with this skill installed
Audit this snippet for a US-targeted Facebook ad: "Our supplement is the #1 doctor-recommended way to lose 20lbs in 30 days, guaranteed. Join 50,000 users today!" No sources provided.
- Read your context and instructions
- Compiled the fact-check & compliance
VERDICT: HOLD
- [s1] "#1 doctor-recommended" — SUPERLATIVE, UNVERIFIED — Fix: Remove or cite a specific study.
- [s1] "lose 20lbs in 30 days" — REGULATED, UNVERIFIED — Health claim lacks data and "results not typical" disclosure. Fix: Remove specific weight/time promise.
- [s1] "guaranteed" — Fix: Remove.
Connects securely to your tools. The creator never sees your data.
About this skill
What it does
The Fact-Check & Compliance Gate acts as a rigorous quality assurance layer for AI-generated or human-written content. It systematically extracts factual, statistical, and comparative claims and audits them against provided source material or external benchmarks. The skill identifies hallucinated figures, unsupported superlatives, and non-compliant marketing language in materials like blog posts, landing pages, and ads.
Why use this skill
Prompting an AI to "be factual" is often insufficient, as LLMs can hallucinate citations or present opinions as facts. This skill is better because it follows a strict multi-step audit protocol: claim extraction, classification (e.g., Regulated, Comparative, Statistical), evidence grading, and risk flagging. It replaces vague feedback with a strict PUBLISH/REVISE/HOLD decision framework, ensuring your content meets legal and editorial standards before it reaches an audience.
Supported content types
- Marketing copy and cold emails (FTC/ASA compliance)
- Scientific or technical blog posts
- Financial and health-related product claims
- PR announcements and press releases
- Landing pages and App Store descriptions
The output
You receive a structured Fact-Check Report containing a final verdict, a list of critical fixes with line-level edit suggestions, specific warnings for unverified claims, and a concise summary of the risk profile.
Verifies claims only against sources the user provides or that are publicly checkable; it cannot confirm private or proprietary data. Opinion and subjective claims are advisory, not hard-verified. It does not run code or perform live web browsing on its own, and compliance flags are guidance, not legal advice.
How to use
- 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
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
This skill has been used 6 times.
Trust & safety
Security scan
Security scan passed 5 months ago
Permissions
Read-only. The skill reads the content under review and any source files the user supplies in order to verify claims. It does not write, execute, or access the network on its own.