fact-check

Verifies generated content claims against external sources and outputs citations with confidence ratings.

Updated Jan 6, 2026
One-click install
npx skills add https://github.com/jwynia/teach --skill fact-check-jwynia
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: fact-check
Source: https://github.com/jwynia/teach/tree/main/.claude/skills/fact-check
Command: npx skills add https://github.com/jwynia/teach --skill fact-check-jwynia

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill systematically verifies claims in generated content by extracting verifiable assertions and checking them against external sources to catch hallucinations.

Core Features & Use Cases

  • Separate verification pass after generation to ensure accuracy and add explicit source citations.
  • Structured claim extraction, categorization, and logging to support audit trails and confidence scoring.
  • Suitable for research synthesis, content generation, and educational materials where factual accuracy is critical.

Quick Start

Run a separate fact-check pass after generating content to verify each claim against external sources and attach citations.

Frequently Asked Questions about fact-check

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify generated content for hallucinations after generation?

To verify generated content for hallucinations, you run a separate fact-check pass that extracts verifiable claims, checks them against external sources, and attaches citations with confidence scores before delivery.

What is a post-generation fact-checking pass and how does it work?

A post-generation fact-checking pass systematically extracts verifiable assertions from text, categorizes them, and confirms them against external sources to produce a structured claims log with source citations and confidence levels.

How do I add source citations and confidence scores to AI-generated text?

You add source citations and confidence scores by applying a verification phase that extracts claims from the generated text, validates them against external sources, and enforces explicit citations in the final output.

Can I use automated claim extraction for research synthesis workflows?

Yes, automated claim extraction is suitable for research synthesis workflows, providing structured categorization and logging to support audit trails while verifying factual accuracy against external sources.

What's the best way to create an audit trail for AI-generated educational materials?

The best way to create an audit trail is running a separate verification pass that outputs a structured log of claims, sources, and confidence scores, ensuring factual accuracy for educational materials.

Why does my AI content generation workflow need a separate verification phase?

Your AI content generation workflow needs a separate verification phase to systematically catch hallucinations by enforcing explicit source citations and confirming claims against external sources before final delivery.