KeelenUse case · OpenAI Codex
Codex comes with your ChatGPT plan. Use a configured workflow.
Connect a Codex or ChatGPT subscription, or an OpenAI API key, to a project. Keelen turns your Requests into planned tasks. It makes changes in isolated runs. Then it returns pull requests and decisions for review. The roadmap keeps moving under your project's capacity and settings.
Choose Claude Code, Codex, GLM, Kimi, or Grok for each project. Use your own supported subscription or API key. Switching engines preserves your roadmap. The Operator plan costs $79 a month. You get 1,000 iterations, 10 projects, 5 concurrent runs, and 5 scheduled projects on the 24/7 loop.
- Connect once with a Codex login. The loop runs on the ChatGPT subscription you already pay for.
- It runs as a standing process. It picks its own next task from your roadmap and keeps going.
- You can set controls per project. They can include red first proof, adversarial review, and a verified clean checkout command. They can also include required checks. A review window can run before gated auto-merge.
- Choose Claude Code, Codex, GLM, Kimi, or Grok for each project. Each one uses your own connected credentials.
Use the subscription you already have
Connect a Codex subscription or API key. Then send one bounded Request. Check its acceptance criteria. Check the branch and review settings too. Provider allowances and billing stay with the provider plan.
Request-to-roadmap progression
Keelen keeps your Requests and a roadmap, and work that fits your settings can keep going. Other agent products may also plan work in the background. Compare the real setup before you choose, and do not assume a tool only handles prompts.
Review and merge controls
A separate adversarial reviewer examines the diff. Red first proof and a clean checkout test run need a verified test command. Gated auto-merge needs the configured checks and a review window. Choose manual merge to keep the final click. Or choose branch-only delivery. It works without a pull request.
Honest limits
Ambiguity comes back as a question. Unclear requests raise one plain English card in your Needs-you queue. Your test suite bounds the autonomy. A repo without tests onboards in an honest no test posture. That lasts until the loop's own PRs add one. Provider limits are handled like weather. Runs back off and retry on a budget, and those retries never count against your iterations.
Recommended plan
Operator: $79 / month
FAQ
Do I need an OpenAI API key, or does my ChatGPT subscription work?
Either works. The recommended default is a Codex login from your ChatGPT subscription. Run codex login and upload the resulting auth file. Then the loop runs with no API charges per token. An OpenAI API key works too; usage then bills to your API account at cost.
Which model does Keelen use for Codex projects?
Codex projects use the configured model through the Codex agent. Engines are chosen per project. A Codex project can sit next to Claude Code, GLM, Kimi, and Grok projects in the same workspace. Each one uses its own credential and bills to its own provider at cost.
Will Keelen eat my ChatGPT usage limits?
You choose how much capacity it gets. Your Keelen tier caps concurrent runs. Scheduling is per project. You can pause a project or your whole account at any time. When OpenAI reports a usage limit during a run, the loop backs off and retries on a budget. Those retries never count against your iterations.
Is Keelen Codex-only?
No. Connect Claude Code, Codex, GLM, Kimi, or Grok on your own supported subscription or API key. Grok accepts a SuperGrok sign in or xAI API key. Pick an engine per project and switch without losing your roadmap or history.