AI agents · DevSecOps

I build AI agents & security tooling —
and ship them end-to-end.

C.Y. Lin (Hayatelin). In 2026 I designed, built, tested and published 10 open-source projects across AI agents, DevSecOps and a full-stack AI product. Try two of them live, right here in your browser.

10 open-source projects 100+ automated tests 0 dependencies in the core tools MIT licensed

Try it live

No install. This runs the actual Python of two of my tools in your browser via Pyodide (WebAssembly). Edit the input and re-run.

promptguard — scan an agent config
skillhub — lint a SKILL.md
agent-config.json
promptguard ./agent
Click "Run promptguard" to scan this config in your browser.

Runtime loads on first run (~a few MB, cached after). Everything executes locally in your tab — nothing is sent anywhere.

Projects

Each ships with docs, tests, CI and semantic-versioned releases. Star counts are live.

How I work

Tested + CI. 100+ automated tests across the code projects, GitHub Actions on each.
Real distribution. Tools ship as a CLI, a GitHub Action, a pre-commit hook, an MCP server and a Python package.
End-to-end ownership. design → build → test → document → release → distribute.
Zero-dependency core. The security & agent tools run on the Python standard library alone.