What problem does it solve? Marketing sites often contain fabricated case studies, stock or AI-generated images captioned as real jobs, reused photos, and invented testimonials. This Skill detects those misrepresentations by fetching and comparing the site's own media assets against their captions, then remediates them into honest equivalents. ## Core Features & Use Cases - Image Fetch-and-Compare: Downloads each case-study image and compares its content to the claimed caption to catch stock/AI images passed off as specific jobs. - Reused-Photo Dedup: Hashes all images (SHA-256) to detect one photo reused as multiple different 'exact' jobs. - Fabricated-Testimonial Detection: Flags pull-quote reviews with city-only attribution and no real author or source. - Honest-Generic Remediation: Reframes fabricated case studies as generic examples, keeps only real reviews, and removes contradicted over-claims like 'no stock photos'. - Use Case: Before launching a service-business website, run this audit to verify every 'our work' photo and testimonial is genuine, then produce a claim-to-honest-replacement diff. ## Quick Start Audit this marketing site's case studies, testimonials, and gallery for fabricated or misrepresented claims and produce a remediation diff.
