Find AI agent skills worth installing — with reviews, rankings, and security audits.
4,274 skills with public reviews, install context, and source data across 24 agents and 784 sources.
Browse by agent
Start where developers already work with AI agents.
The index should feel familiar to builders scanning the environments they already use.
Browse by category
Compare AI skills by job to be done.
Categories should compress the field quickly without hiding the underlying signal.
Live catalog rankings
Proof-led ordering for the skills people install most.
Rankings combine install activity, sync freshness, and source-linked trust signals so the list stays useful for comparison.
About
About SkillJury
SkillJury indexes 4,274 skills from 784 sources across 24 AI agents including Claude Code, Cursor, Windsurf, Codex, and Cline. Each listing includes security audit signals from Socket.dev and Snyk.
Community reviews are separated from catalog signals like weekly installs, GitHub stars, and source metadata. This makes it possible to shortlist skills by operational data while evaluating them through independent reviews.
FAQ
Questions developers ask first.
What is SkillJury?
SkillJury is the public review layer for AI agent skills, combining live catalog data with review slots, source pages, compatibility context, and security-audit signals.
What do the leaderboard tabs mean?
All Time ranks skills by current weekly install volume, Trending favors recently synced skills with install momentum, and Hot surfaces the newest active entries in the catalog.
Are leaderboard rankings the same as public reviews?
No. Install counts and source metadata help developers shortlist skills, but community reviews remain separate so operational telemetry never pretends to be a verdict.
Can I browse beyond the homepage?
Yes. SkillJury has dedicated pages for individual skills, agents, categories, sources, and public review archives so developers can go from discovery into deeper evaluation.