Finding the right plugin or skill for your AI coding agent currently means knowing the exact GitHub repo name, or scrolling through endless lists. A new open source project called SkillsGate is trying to fix that by indexing over 45,000 skills from GitHub repositories into a single searchable marketplace.
The project covers skills for Claude Code, Cursor, Windsurf, and other AI coding agents. Each skill has been enriched with LLM-generated metadata and indexed using vector embeddings - essentially, the search understands what you want to do rather than requiring exact keyword matches. Search "help me write better commit messages" and it finds relevant skills, even if none of them use that exact phrasing.
The practical value here is discovery. The AI coding agent space has exploded with community-built extensions, custom commands, and workflow automations, but finding them is a mess. Most live in scattered GitHub repos with inconsistent documentation. SkillsGate pulls them into one place where you can browse, search, and install.
At 45,000+ indexed skills, the catalog is large enough to be useful but also raises obvious quality questions. Automated LLM-generated metadata can miss nuance, and semantic search over that many entries will surface some irrelevant results. The project is open source, though, so the community can improve both the indexing and the search quality over time.
For anyone building workflows with Claude Code or Cursor, this is worth bookmarking. The AI agent skills ecosystem is growing faster than any single developer can track, and centralized discovery tools like this fill a real gap.