What problem does it solve?
Many blog posts contain numeric claims, percentages, dollar figures, or named attributions without clear verification; this Skill identifies those claims and checks whether cited sources actually contain the referenced data so authors and editors can trust published content.
Core Features & Use Cases
- Claim extraction: Scans markdown blog posts to extract sentences containing numbers, percentages, dollar amounts, multipliers, and named attributions along with their location and any cited URL.
- Source verification: Fetches cited URLs via WebFetch, searches page content for exact or contextual matches, and assigns a confidence score (VERIFIED, PARAPHRASE, WEAK, NOT FOUND, UNVERIFIED).
- Editorial guidance: Flags uncited claims with suggested search queries, detects paywalls or PDF sources for manual review, and enforces a 10-URL-per-run rate limit to avoid overloading sources.
- Use Case: An editor preparing a data-forward blog post runs the Skill to produce a verification table and recommended actions before publication.
Quick Start
Run blog-factcheck on path/to/post.md to extract claims, fetch cited sources up to ten URLs, and generate a verification report.
