Turn tracked tickets into pull requests

Use an agent to pick up issues or tickets, make the code change in an isolated pod, and open a PR for review.

Done with Claude CodeCodex· Advanced

What they did
A team ran Optio on Kubernetes and handed it work from GitHub Issues, Linear, or Jira. It created or reused a long-lived pod for the repo, added a git worktree, and let Claude Code or Codex make the change there. A PR watcher kept checking CI and review status, then sent the agent back in with failure details, merge conflicts, or reviewer comments. A separate review agent also took part.
How it went
The workflow is designed to carry a task from ticket to merged PR, then squash-merge once CI passes and review is complete. After that, it closes the linked issue and logs cost. The README does not give any success-rate figures or benchmark results.
Tips
This setup depends on Kubernetes v1.33 or later, plus Helm, Postgres, and Redis. Auto-resume stops after the limit set by OPTIO_MAX_AUTO_RESUMES, so some PRs can still need a person to step in.

Try it yourself with Codex

I have a repository at [owner/repo] and a set of tickets in [GitHub Issues/Linear/Jira]. Find the [number] smallest open items tagged [label]. For each one, use a separate branch or worktree, make the code change, run the relevant tests, and prepare a draft PR summary. Stop before pushing, opening the PR, deleting anything, or submitting any final action, and show me the diff plus the draft PR text for approval. Only after I approve, continue with the next step.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on github.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.