Let a background agent watch tasks and open PRs
A team keeps a server-based agent running so Cursor can monitor work and turn finished tasks into pull requests.
- What they did
- They host a Claude Code server on a VPS and use the Caliber AI setup repository to prepare their environment. That repository builds the MCP settings and agent instructions for both the codebase and Cursor, and they reload everything with a single command. After that, a nonstop script tracks work items and sends Cursor in to open pull requests.
- How it went
- They used this setup to ship several internal utilities, including a document summarizer and a data pipeline. The comment did not include any numbers, schedule, or review details.
- Tips
- This is most useful if you are comfortable maintaining extra infrastructure, since it depends on a VPS, generated configuration, and a sync step. Make the agent stop for approval before any pull request is created or submitted.
Try it yourself with Cursor
I want a long-running agent workflow for [project]. Keep checking [task source] for items that match [rules]. When you find one, use Cursor to implement it, then pause and ask for my approval before creating or submitting any pull request. If you need more context, ask before changing files. After each task, give me a short progress note and what you’re waiting on.
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Use Cursor to inspect pod logs and fix compose files
A DevOps engineer used Cursor to read production pod logs through Kubernetes MCP and to repair a broken Docker Compose setup.
Flag coding conflicts before agents edit
They set up a local layer that makes coding agents announce planned changes so overlapping architecture work gets caught early.
Have agents co-build a reference system with role splits
A solo founder divided an AI-assisted build into roles and shipped a 120-model mental-models framework with an MCP server.
Use Cursor to spot prototype gaps and draft a build plan
A founder used Cursor to inspect a trip-planning app, surface missing pieces, sketch screens, and then build a feature with parallel agents.
Hand small repo fixes to Cursor agents
A developer used Cursor agents to chip away at a daily batch of repo maintenance, bug fixes, and minor features.
Assign PRs and bug fixes across Cursor sub-agents
A Cursor coordinator tracked incoming pull requests and live bugs, sent sub-agents after each issue, and kept the user updated along the way.