Coordinate sub-agents to monitor PRs and fix prod bugs
A Cursor Projects coordinator agent delegates to sub-agents, watches new PRs and production bugs, spins up agents to fix issues, and surfaces questions to the user.
The author had Devin build and deploy an app, finding it quick on PR feedback but weak on legacy monorepos and costly on broad requests.
The author had Devin build and deploy an app, finding it quick on PR feedback but weak on legacy monorepos and costly on broad requests.
The author subscribed to Devin's Core plan and gave it a blank GitHub repo, asking for a City Explorer dashboard showing images, weather and news for a city. Devin ran in its own cloud VM. It scaffolded a FastAPI backend and React frontend, set up .env files and routes, added placeholder integrations, tested locally, opened a PR and deployed. The author only approved or clarified one or two prompts.
A working, deployed app took about 90 minutes from an empty repo. Devin is weaker on older monorepos and bespoke infrastructure, and broad requests burn more compute units.
Devin bills in Agent Compute Units, and idle time and thinking both use them. Keep requests narrowly scoped and kill idle sessions to hold down cost.
Build and deploy a small app that [describe what the app does] in a new repo. Keep the scope narrow, start with the simplest working version, and tell me your plan and rough effort before you begin. Finish by giving me a live URL and a short list of what works and what doesn't. Pause and ask me before deploying or making any change that could cost money.
A Cursor Projects coordinator agent delegates to sub-agents, watches new PRs and production bugs, spins up agents to fix issues, and surfaces questions to the user.
Ran a Cursor agent continuously for 24 hours to build a project management tool; the app fully works but is not production ready.
The author had ChatGPT take the Session Buddy extension source from Chrome, port it to Safari and install it, taking about 20 minutes.