AI Skills Manager

Alexey Glukharev, Takahito Yoneda, Mykola Kondratiuk, and the team at Zuna Labs looked at the AI coding agent landscape and realized something obvious that somehow nobody had solved yet: every single one of these agents—Claude Code, Cursor, Windsurf, Copilot, Gemini CLI, Goose, OpenAI Codex, OpenCode, Kilo Code, Trae, Antigravity—works differently. They all have their own skill systems, their own prompt formats, their own ways of getting better at your specific workflows. And if you're like most developers, you're probably using two or three of them at once. That means you're managing three different sets of skills manually, and that's a nightmare.

So they built AI Skills Manager—one place to install, manage, and share skills across every major coding agent. Eleven of them. All in one app. It's open source under MIT, no accounts required, works on Windows and Linux, and the source is right there on GitHub for anyone who wants to peek under the hood or contribute. They launched on Product Hunt and caught fire because this is a problem every AI-powered developer actually has but nobody had bothered to solve.

The beauty is in the simplicity—no login, no cloud, just a local manager that keeps your skills organized and portable. Want to share that killer prompt you built for debugging? Just export it and anyone running AI Skills Manager can pick it up and use it in their agent of choice. As AI coding agents keep proliferating, something like this becomes essential infrastructure. The tool that manages all your AI superpowers.