fact-checking

Extract claims from documents, verify evidence, and assign verdicts.

2|Updated Feb 4, 2026
One-click install
npx skills add https://github.com/PaulKinlan/docker-agent-test --skill fact-checking-paulkinlan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: fact-checking
Source: https://github.com/PaulKinlan/docker-agent-test/tree/main/config/skills/editor/fact-checking
Command: npx skills add https://github.com/PaulKinlan/docker-agent-test --skill fact-checking-paulkinlan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill systematically verifies factual claims within a document, ensuring accuracy and reliability before publication or decision-making.

Core Features & Use Cases

  • Claim Extraction: Identifies and lists all verifiable factual statements in a document.
  • Evidence Verification: Searches available sources to confirm or contradict each claim.
  • Verdict Assignment: Assigns a status (Confirmed, Unconfirmed, Contradicted, Modified, Opinion, Projection) to each claim.
  • Report Generation: Produces a detailed report with summary statistics and an annotated version of the document.
  • Use Case: Before publishing a research paper, use this Skill to verify all statistical data, experimental results, and cited facts against the source material, flagging any discrepancies.

Quick Start

Use the fact-checking skill to verify all claims in the document located at /home/shared/research_paper.md.

Frequently Asked Questions about fact-checking

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

FAQPage Schema
How do I verify factual claims in a document before publication?

To verify factual claims in a document, you need a systematic fact-checking process that extracts statements, searches for supporting evidence, and assigns a verdict to ensure accuracy before publication.

What is the best way to extract and check supporting evidence for research paper statistics?

The best way to check supporting evidence for research paper statistics is to use automated fact-checking that identifies verifiable statements and searches available sources to confirm or contradict each claim.

Can I generate an annotated document highlighting unconfirmed or contradicted claims?

Yes, you can generate an annotated document highlighting unconfirmed or contradicted claims by assigning a verification status to each extracted statement and flagging discrepancies in the final report.

Does document analysis for accuracy require external sources to assign a verdict?

Yes, document analysis for accuracy requires searching external sources to assign a verdict, categorizing each extracted claim as confirmed, unconfirmed, contradicted, modified, opinion, or projection.

What types of verdicts are assigned during evidence verification of a text?

During evidence verification, verdicts assigned to extracted claims include confirmed, unconfirmed, contradicted, modified, opinion, and projection, detailing the accuracy of the document.

What are the limitations of automated fact-checking for opinion or projection statements?

The limitation of automated fact-checking for opinion or projection statements is that they cannot be definitively verified against sources, resulting in an unconfirmed or opinion verdict rather than confirmed.