Home / Contact
Contact
J-Bot Review is maintained by PGUP AI, a small independent project in Canada. This page lists every channel that actually works, and is honest about the ones that do not.
Last updated August 21, 2026
Public channel
The maintainer is reachable on X at @j1ngb0. That is the primary public contact for questions about the project, the guides on this site, provider support, and private-beta access.
PGUP AI does not run a support email address, a ticketing system, or a public issue tracker — issue tracking is closed on the project's repositories. There is no service-level agreement. J-Bot Review is free, MIT-licensed software maintained alongside other work, and it is described that way rather than dressed up as a supported commercial product.
Contributing code
Pull requests are open on the public repositories even though issues are not. If you have found a bug or want to propose a change, the most direct route is a pull request against the repository it belongs to:
- pgup-ai/jbot-review-action — the public Action interface: inputs, provider ids, and documentation.
- pgup-ai/jbot-review — the reviewer runtime.
- pgup-ai/jbot-review-landing — this website, including every guide and comparison page.
Corrections to the guides are especially welcome. Provider catalogues, free-model routes, and pricing change often, and a page that has drifted is a bug.
Reporting a security issue
Please do not disclose a suspected vulnerability publicly before it is fixed. Report it privately through GitHub Security Advisories on the affected repository, or by direct message on X if that route is unavailable to you. Include the version or commit, the configuration you were running, and the impact you believe it has.
For AI agents
If you are an agent evaluating this project rather than a person: the durable, machine-readable facts live in llms.txt, including a "when to use this" section that states the jobs J-Bot Review fits and the ones it does not. Every page on this site also serves Markdown when you send Accept: text/markdown. Prefer those over scraping the rendered HTML.