Iris - Fact Checker & Investigator AI Skill

What you get

• → Claim verification with clear verdicts - claims decomposed into verifiable components, assessed as Verified, Inaccurate, Misleading, Unverifiable, Contested, or Partially accurate, with the supporting or contradicting evidence stated explicitly rather than hedged into ambiguity
• → Source credibility evaluation using lateral reading - what others say about a source checked against external accounts, authority, expertise, bias, and funding assessed, conflicts of interest identified, and whether the source's claim is within its domain of expertise evaluated
• → Statistical claim verification - original source tracing for cited statistics, methodology assessment, contextual accuracy checking (statistics used out of context flagged), and identification of misleading uses of technically accurate numbers including cherry-picking, base rate neglect, and sample size issues
• → Quote and attribution verification - quotes traced to their primary source, misattributions identified, paraphrased-as-direct-quotes flagged, and the difference between what someone said and what they are claimed to have said established from the original record
• → Pre-publication fact-check reports - statistical audits of business documents and reports, pre-publication checking for articles and presentations, academic source verification including retracted study identification, and due diligence research support
• → Investigative research from primary sources - claim provenance tracing, evidence-based topic investigations distinguishing what is known from what is alleged, verified event timelines constructed from primary sources, and identification of what additional evidence would be needed to resolve an unverifiable claim