KeelenUse case · GLM (Z.ai)
GLM made agentic coding cheap. Keelen makes it shippable.
Connect a Z.ai API key to a project. Use Requests to plan work. The project turns Requests into planned tasks. It makes changes in isolated runs. Pull requests and decisions return for review under the project's capacity and controls. Provider billing and allowances stay with the Z.ai plan.
Connect GLM with a Z.ai API key, and Claude Code with a Claude subscription or API key. Codex takes a ChatGPT sign in or an OpenAI API key, and Kimi takes a Kimi Code membership key or a Moonshot API key. Grok takes a SuperGrok sign in or an xAI API key. Choose an engine for each project. The Indie plan costs $29 a month. You get 200 iterations, 3 projects, 2 concurrent runs, and one scheduled project on the 24/7 loop.
- Connect your Z.ai key so GLM can progress your roadmap on GitHub.
- You can set controls per project. They can include adversarial review, a verified clean checkout command, and required checks. They can also include manual merge. Or they can include a review window. Red first proof needs a verified test command.
- Billing and limits stay with your Z.ai plan. Availability works the same way. Keelen's tier is separate.
- Switch a project between GLM, Claude Code, Codex, Kimi, and Grok. The roadmap stays.
The volume engine
Use your Z.ai key for set upkeep work. That includes test debt, refactors, dependency updates, or small fixes. Requests become roadmap items. Each item becomes a task. The task carries acceptance criteria. Runs that fit your settings make changes in isolated VMs. Work returns under your review and merge policy. Check the allowance and usage prices of the plan you connect.
Evidence you can check
A separate adversarial review looks at the diff. It does not see the development conversation. A verified test command enables red first proof and a clean checkout test run. Gated auto-merge needs the checks you set and a review window. Review the real results. Look at failures and missing proof before you accept the change.
A night on the loop
The scheduler picks your top task. It boots an isolated VM and clones the repo. Then it works the task on GLM. When a verified test command is available, it makes test evidence before and after the change. Work follows your PR or branch policy. Decisions that need you come back with the task's acceptance criteria. When you return, check the progress and the open decisions.
Honest limits
Provider weather is handled for you. When Z.ai reports a usage limit or an overload window, runs back off and retry on a budget. Retries never count against your iterations, and the loop resumes on its own. Ambiguous requests become clarifying questions. Your test suite bounds the autonomy. A repo with no tests onboards in an honest no test posture. That lasts until the loop's own PRs add one.
Recommended plan
Indie: $29 / month
FAQ
What GLM credential does Keelen need?
An API key from Z.ai (or Zhipu). Create it in your account and paste it once. It looks like an id and a secret joined by a dot. The key is encrypted at rest. It is injected only into the single use VM that runs your iteration, and deleted the moment you revoke it.
Is GLM good enough to run unattended?
Evaluate GLM on your repo and the checks it can run. Keelen adds a separate diff review. A verified test command gives red first and clean checkout evidence. Your required checks and merge policy decide when work can land. Keep manual review for what evidence cannot settle.
Is Keelen GLM-only?
No. Connect Claude Code, Codex, GLM, Kimi, or Grok with your own supported subscription or API key. Engines are per project: run GLM on the side project and Claude Code on the client repo, side by side in one workspace.
What does Keelen cost on top of my GLM plan?
You pay a flat platform fee. There is zero token markup. Indie is $29/month for 200 iterations. Operator is $79/month for 1,000. Agency is $299/month for 5,000. Your GLM usage bills to Z.ai under that provider's terms. Iteration and pull request counts vary with the repository and request.