AI-Implement label (Linear) or set AI-Implement Status to Ready (Jira) on an issue, and within minutes Claude Code checks out the repo, implements the ticket, opens a pull request, and posts a gap analysis comparing what was built to what was asked. Your team reviews in the tools they already use — nothing changes about how you ship.
Quick Start
Get from zero to your first AI-generated PR in under 30 minutes.
How It Works
Understand the full lifecycle from ticket to merged PR.
Setup Guide
Step-by-step instructions for connecting your ticketing system, GitHub, and your runner.
Customize Prompts
Tailor the WORKFLOW.md and PLANNING.md templates for your stack.
How it works in 60 seconds
1
Mark an issue
- Linear
- Jira
Add the
AI-Implement label to any Linear issue. Write a clear, focused ticket — that’s your prompt.2
Claude implements it
The orchestrator picks up the issue, dispatches a GitHub Actions workflow in your target repo, and Claude Code reads your codebase and implements the change.
3
Review the PR
A pull request appears in your repo with a gap analysis comment showing what was implemented and what (if anything) still needs attention.
4
Iterate with /ai-implement
Comment
/ai-implement on the PR to send Claude back in for a gap-fill pass on the same branch.What makes it different
AI-Implement is built for teams, not individual developers. The ticket is the prompt, and your existing ticket backlog is the context. Every run produces a PR, a gap analysis, and a ticket-system status update — so reviewers always know exactly what was attempted and why.Works in your CI
Runs inside GitHub Actions with your own secrets. Nothing leaves your infrastructure.
Multiple repos & orgs
One orchestrator, many repos across many GitHub organizations.
AWS Bedrock support
Use Anthropic’s API, Claude OAuth, or AWS Bedrock — configured per repo.