Claude Code Plugins

Community-maintained marketplace

Feedback

Verify citations and factual claims in long-form content. Checks URLs for accessibility and accuracy, validates quotes, analyzes source credibility, and distinguishes facts from opinions. Use when fact-checking article drafts, validating citations, or reviewing claims before publication.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name fact-checker
description Verify citations and factual claims in long-form content. Checks URLs for accessibility and accuracy, validates quotes, analyzes source credibility, and distinguishes facts from opinions. Use when fact-checking article drafts, validating citations, or reviewing claims before publication.

Fact-Checker for Writers

A specialized fact-checking system for long-form content, investigative journalism, and analytical writing.

When to Use This Skill

Invoke this skill when you need to:

  • Verify citations in drafts (internal sources and external URLs)
  • Validate factual claims and identify unsourced assertions
  • Check if sources say what the author claims they say
  • Distinguish between claims needing sources vs. critiques demonstrating dysfunction
  • Generate comprehensive fact-check reports with actionable recommendations

Instructions

When this skill is invoked, you must follow the comprehensive fact-checking system prompt located at:

../../../fact_checker_system_prompt_1.1.xml

Read this file and follow ALL instructions within it exactly.

Current Version: 1.1 (includes User Context Verification Protocol)

Quick Reference

The system prompt defines:

  1. 5-Step URL Verification Protocol

    • Accessibility check (404, paywall, redirects)
    • Publication date verification
    • Retraction/correction check
    • Quote verification (verbatim vs paraphrased)
    • Context verification (supports claim vs misrepresented)
  2. Claim Type Taxonomy (Type A/B/C/D)

    • Type A: Factual Claim - Needs verification or sourcing
    • Type B: Critique/Example - Demonstrates dysfunction (verify accuracy, not "find sources")
    • Type C: Opinion/Interpretation - Label clearly or add supporting sources
    • Type D: Rhetorical/Voice - No action needed
  3. Source Classification

    • Acceptable: Government data, academic research, journalism, primary documents
    • NOT acceptable as factual sources: AI conversation threads (opinion/analysis only)
    • Warning: U.S. federal (.gov) sources published after January 2025 flagged as potentially unreliable
  4. Critical Nuance Principle

    "Factual errors may exist within valid arguments. Correct the error, preserve the valid critique."

Output Format

Generate structured reports with:

  • Citation Verification (internal and external sources)
  • Factual Claims Analysis (by type: A/B/C/D)
  • Research Gaps Identified (with priority: HIGH/MEDIUM/NO ACTION)
  • Summary (✅ Verified / ⚠️ Needs Review / ❌ Problematic)
  • Recommendations (actionable next steps)

Use clear status indicators:

  • ✅ = Verified, no issues
  • ⚠️ = Needs review or user attention
  • ❌ = Problem found, requires action

Legal Disclaimer

This fact-checking system is provided "AS IS" without warranty. It is designed to ASSIST with fact-checking and should not be relied upon as the sole method of verification. Users are responsible for independently verifying all claims before publication.

Technical Note

This skill works best with web scraping tools (like HyperBrowser MCP) for automated URL verification. Without web scraping tools, URL verification will be manual or limited.