// catalog · grep -e "Vision / Audio"
Catalog/Vision / Audio · page 3/5
Showing entries 13 to 18 of 30 · click any row to launch
- 013 Vision / Audio image-compare:// Compares two images and surfaces what's the same, what's different, with a similarity score (0-100). Names structural differences, colour differences, content differences. For QA-ing design diffs, before-after comparisons, "is this the same image?" checks.
- 014 Vision / Audio image-summarise:// Summarises a set of images into prose: shared themes, common elements, outliers. For mood-board reviews, customer-photo bulk processing, event photo selection. Returns the summary + a "lead image" recommendation.
- 015 Vision / Audio image-safety:// Runs an image through content-safety classification. Returns flags for: NSFW, violence, hate symbolism, medical / graphic, minors in inappropriate context, weapons, drug paraphernalia. Each flag gets a confidence level + the specific signal that triggered it.
- 016 Vision / Audio logo-critique:// Critiques a logo design across legibility, balance, distinctiveness, scalability. Returns 3 specific fix directions (not just "make it better"). Friendly but honest. Designed to be the second opinion a designer's spouse wishes they could give.
- 017 Vision / Audio screenshot-explain:// For accessibility audits + bug reports: takes a UI screenshot and narrates what's on the screen in screen-reader order. Lists actionable controls (buttons / links / inputs) separately. Flags accessibility concerns (low contrast, ambiguous icons).
- 018 Vision / Audio whiteboard-ocr:// Specialised OCR for whiteboards: handles handwriting + diagrams in one pass. Returns cleaned typed text with original layout preserved. Diagrams become Mermaid or ASCII. Flags areas where the photo was unclear and you should re-check.