VisionFlow Match

Each tool targets a distinct matching approach: template-match for same-scale pixel matching, homography for scale/rotation/perspective, and feature-match for raw keypoint correspondence. The descriptions explicitly cross-reference each other to guide selection, though feature-match and homography share the same ORB/SIFT+matcher mechanism and could occasionally be confused.

Naming is consistent in kebab-case, but the pattern is mixed: feature-match and template-match use noun-match, while homography is a bare concept noun with no verb or action. Readable but not a predictable convention throughout.

Three tools is slightly thin but well-scoped for a focused image-matching service, and each tool earns its place by covering a genuinely different matching strategy. It sits at the low end of the comfortable range without feeling padded or truncated.

The surface covers the main pattern-localization strategies (exact template, geometric homography, raw feature correspondence), which is the core domain. Minor gaps exist around batch processing or keypoint/visualization output, but agents can work around these.