Hand migrations and bug-fix tickets to a coding agent
Reviewers gave Devin migration, data engineering, bug fix and refactoring tickets on a Next.js monorepo and Python pipeline over three weeks; it did best on migrations and framework upgrades.
Done withDevin
- What they did
- Reviewers spent three weeks assigning Devin migration, data engineering, bug fix and refactoring tickets on two production codebases, a Next.js monorepo and a Python data pipeline. Tickets came in through Slack, Linear or Jira. Devin planned the work, wrote code in its own cloud environment, ran tests and opened pull requests. Reviewers could step in on its plan and leave PR comments for it to act on.
- How it went
- It handled repetitive, well-defined work well, especially migrations, and mostly got review comments right the first time. Vague requirements gave mediocre to unusable results, and it could spend about 20 minutes on a wrong path before anyone noticed.
- Worth knowing
- Cost was hard to predict: two similar migration tasks used 5 and 28 ACUs, at $2.25 per ACU on Core. Good playbooks also take hours to write.
Try it yourself with Devin
I'm giving you a batch of tickets for my repo [repo URL]: [list 3 to 5 tickets such as a framework upgrade, a data migration, a bug fix and a small refactor]. Work on them one at a time, each on its own branch with tests passing, and open one pull request per ticket. For each, write a short note on what you changed and any risks, and stop there so I can review.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: popularaitools.ai · Undated
Related use cases
Hand narrow code changes to Devin for pull requests
A team used Devin for a month on small fixes like chart formatting and API edge cases, with tests and PRs; a complex full-stack task went poorly.
Rewrite a README after a rearchitecture
Devin scanned a rearchitected codebase and delivered a new README as a PR with accurate setup and API details, but struggled with design rationale and purpose.
Split a React Native to Swift migration across agents
Cognition's Jared Zoneraich uses Devin fan-out so child agents work in parallel on a migration split into six pieces.
Build a C compiler with a team of parallel agents
Anthropic's engineering write-up on tasking Opus 4.6 agent teams with building a C compiler.
Build a homepage from Figma designs via MCP
The author connected Figma to Cursor over MCP to build a new homepage for a font archive, with great results after Lovable disappointed.
Run Codex with low-touch planning and port reservation
A commenter has Codex execute mostly on its own with human planning and review, including a portbroker that reserves and injects a port.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.