Why this exists
AI Skills are becoming fragmented across platform stores, GitHub repositories, agent ecosystems, and community posts. When supply is fragmented, users need a cross-platform decision layer instead of another popularity list.
Product logic
- Start from the task instead of showing a wall of skills.
- Evaluate fit with evidence, output samples, failure signals, and safety notes.
- Let users contribute real tasks, skill links, and usage results over time.
What I want to validate
The core question is not whether people browse rankings. It is whether users search, compare, and submit cases when they do not know which AI Skill to use.