Fact-Checking & Misinformation Detection
Verifies claims, evaluates source credibility, and detects disinformation using structured evidence ledgers and lateral reading techniques.
The Fact-Checking skill equips Claude with rigorous, transparent verification workflows to analyze claims, articles, URLs, images, and social posts across multiple languages. By implementing claim decomposition, evidence-ledger tracking, manipulation scanning, and lateral reading, this skill enables nuanced source evaluation and narrative prebunking. It supports both quick checks and detailed multi-claim investigations in areas like EU policy, science, health, and geopolitics, complete with tools to render responsive HTML Fact-Check Cards and share-safe neutral summaries.
主な機能
01Multi-mode verification pipeline supporting quick checks, standard investigations, source comparisons, and prebunking
02Automated output formatting for neutral share-safe summaries and validated HTML fact-check cards
0375 GitHub stars
04Evidence-ledger based claim decomposition and structured source reliability tiering
05Multilingual support with dedicated regional context for English, Bulgarian, Russian, and EU policy
06Manipulation scanning to flag emotional framing, statistical abuse, false context, and AI markers
ユースケース
01Verifying breaking news, viral social media claims, quotes, or screenshots before sharing or publishing
02Prebunking emerging disinformation narratives and generating educational media-literacy briefings
03Comparing two conflicting news articles or sources to identify contradictions and relative reliability