Thanks! We've moved the disclaimer up the page :)

It's GitHub, using GitHub Pages for docs,

  https://github.github.com/gh-aw/ 

served as GitHub Pages from

  https://github.com/github/gh-aw

It uses Copilot, Claude Code or OpenAI Codex. Custom engines/coding agents also possible.

These contain links to sample outputs:

* Quality/Hygiene: https://github.github.com/gh-aw/blog/2026-01-13-meet-the-wor...

* Documentation: https://github.github.com/gh-aw/blog/2026-01-13-meet-the-wor...

* Code improvement: https://github.github.com/gh-aw/blog/2026-01-13-meet-the-wor...

* Refactoring: https://github.github.com/gh-aw/blog/2026-01-13-meet-the-wor...

You are 100% in control.

Generation of the lock file is deterministic.

See here for information about determinism https://github.github.com/gh-aw/reference/faq/#determinism

This firewall is enabled by default

> Separating the LLM invocation from the apply step is a great idea

Thanks, yes, this is crucial.

Thanks, this has been changed (no use of local token) and there are now extra confirmations too.

Is this thing production ready?

Hey, sorry, yes the better link is https://github.github.com/gh-aw/ but we had a redirect set to https://github.github.io/gh-aw/

Both work and we've fixed the redirect now, thanks