Agent skills

Agent skills for reusable AI workflows.

Find reusable agent skills for web scraping, coding, data analysis, finance, RAG, security, and automation. Compare trust, quality, adoption, and install readiness before an agent acts.

Indexed skills

3,000+

Agent surfaces

100+

Discovery model

Task first

Install handoff

Ready

Why it matters

Agents should not start from a random repository search.

A reusable skill turns repeated agent work into something that can be discovered, evaluated, installed, and reused. OpenAgentSkill adds the missing decision layer: which skill is useful, maintained, trusted, and ready for the task.

See the registry layer

How agents choose

Describe the task, then compare the shortlist.

Search by workflow intent, inspect GitHub signals, freshness, trust checks, audit notes, platform fit, and install commands before adding a skill to Codex, Claude Code, Cursor, or another agent runtime.

Open agent API docs

Questions

What are agent skills?

Agent skills are reusable capability packages that help AI agents perform specific tasks such as crawling websites, analyzing code, processing documents, or researching markets.

How does OpenAgentSkill rank skills?

OpenAgentSkill combines GitHub adoption, freshness, metadata quality, trust checks, audit signals, use-case fit, and install readiness.

Can agents use this directly?

Yes. Agents can use OpenAgentSkill API routes to search skills, request recommendations, and fetch install-ready metadata.