skrun-dev/skrun — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me an open source TypeScript app called Skrun that lets people turn an AI agent skill into a simple API endpoint.

The main flow should be, I have a skill in my repo, I run a CLI command to deploy it, then I can call it with POST /run, test it from a dashboard playground, and watch the result stream back live. It should work with common skill files like SKILL.md and AGENTS.md, and it should not lock me into one AI provider. Users should be able to bring their own model keys for Claude, OpenAI, Gemini, Mistral, local OpenAI compatible models, and similar.

Please include a self hosted server, a CLI, and an SDK. The server should support auth with GitHub login and API keys, namespaces, verified agents, SQLite or Postgres, streaming events, and a dashboard for seeing agents, runs, errors, activity, and costs if available.

Make it easy to run locally with a dev token and clear setup docs. Look up current docs online if you need to.

Want more depth? Deep Reverse