Gitlawb/openclaude-skills — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me a small TypeScript repo for an OpenClaude skill hub. I want it to be the source of truth for installable skills that the openclaude CLI can read from a generated registry.json file, with no server or API. Each skill should live in its own folder under skills with a markdown SKILL.md file and an optional README, and the project should rebuild the registry from those markdown files automatically.

Please include validation so malformed or unsafe skill content gets rejected, plus trust levels like official, verified, community, and deprecated, with maintainer controlled promotion and a revocations list for compromised or withdrawn versions. Seed it with the example skills from the README, like PR review, debugging, security audit, test writer, frontend implementation, CI fix, database review, git conflict resolve, provider debug, and CodeQL fix, so the registry feels real. Add simple scripts for validating one skill and rebuilding the full registry, and make the whole thing ready to work with the OpenClaude CLI. Look up current docs online if you need to.

Want more depth? Deep Reverse