Your muscle memory with AI agents, the prompts you've refined, the context your codebase always needs its probably rotting in a folder / brain. I sat with that problem long enough to get annoyed, then built something.

⚙️ skillpod - a CLI that puts you in charge of your own skill registry. Create skills, update them, push them, sync across machines, install into any agent you work with. Claude Code, Codex, Cursor - it handles the wiring so you don't have to.

Install it. <Favorite Package Manager> <install / add> -g skillpod

~6,000 lines of code. ~100 test cases. Built entirely with Claude Code and Codex. Shipped to npm.

The thing I didn't expect: building this forced me to articulate what a "skill" actually is. It's not a markdown file with instructions. It's compressed judgment. It's the difference between an agent that kinda works and one that works the way you want it to work.

The people who figure out how to preserve and version that judgment across machines, across projects, across agents - are going to move at a different speed than everyone else. skillpod is how I'm doing it.

#AgenticAI #ClaudeCode #DeveloperTools #AIEngineering #OpenSource