Claim Investigation

Breaks down viral claims and social media posts into verifiable atomic components, then systematically checks each piece against official sources and procedural context. You'll decompose claims like "X refused to do Y because Z" into separate assertions about entities, events, causation, and narrative framing, then investigate each independently. The output structure separates verified facts from disputed elements and narrative interpretation, with confidence levels for each claim. It's especially useful when you encounter politically charged assertions that mix provable facts with unverifiable motivations. The approach forces you to find what can actually be confirmed versus what's speculation, and it documents procedural context that often gets lost in hot takes.

Install to Claude Code

npx -y skills add jwynia/agent-skills --skill claim-investigation --agent claude-code