Skip to content

How it works ​

1
Fetch

Pull tagged tasks from your provider

2
Clarify

Ask questions when descriptions are unclear

3
Implement

AI agent writes code on a fresh branch

4
Ship

Commit, push, open PR, move to review

5
Agent review

AI reviews the PR diff on tagged tasks

6
Code review

Resolve human PR comments automatically

7
Merge

Squash-merge accepted PRs via gh

Step by step ​

  1. Fetch — pulls all tasks tagged with your configured tag from your task provider
  2. Filter — skips done/cancelled tasks, tasks that already have a branch, and tasks blocked by other open tasks
  3. Clarify — in smart mode, asks the AI if the task description is clear enough; if not, posts a question as a comment (prefixed with [aidev]) and marks the task pending
  4. Wait — pending tasks are re-checked on the next run; if a human replied or the trigger word is found, implementation proceeds with the conversation as context
  5. Implement — checks out a fresh branch (or reuses an existing one), runs your configured AI agent(s), falls back to the next agent if one fails
  6. Ship — commits all changes, pushes the branch, posts a comment with the branch name and a PR link, moves the task to your "in review" status
  7. Agent review — for in-review tasks tagged with agent review, fetches the PR diff via gh, runs an AI review, and posts inline comments (or an approving review when no issues are found)
  8. Code review — checks tasks already in review for unresolved human PR review comments; if found, runs an AI agent to fix code or reply to discussion threads
  9. Merge — tasks tagged accepted are squash-merged via gh (see Auto-merge)

At a glance ​

Task  →  AI implements  →  git push  →  "in review"  →  agent review  →  resolve PR comments  →  accepted  →  merge

Released under the MIT License.