Jules (Google Labs): async agent that resolves GitHub issues autonomously
In one sentence Google Labs launches Jules: assign a GitHub issue, Jules clones the repo in an isolated VM, implements the fix, runs tests, and opens a PR. First async coding agent from a major player natively integrated into the GitHub workflow.
Imagine being able to assign a bug or feature to a colleague who works completely autonomously, only notifying you when done with the pull request ready. Google's Jules works exactly like this, but instead of the colleague there is an AI agent.
The workflow is simple: go to a GitHub issue, assign the issue to Jules, and then you can go do other things. Jules, meanwhile, clones your repository in an isolated virtual environment, reads the issue description, analyzes the existing code to understand where to intervene, implements the change, runs the test suite to verify everything works, and finally opens a pull request with the changes ready for your review.
The difference from coding agents that run in the IDE (like Cursor or Copilot Workspace) is that Jules works completely asynchronously on remote servers: you do not need to keep your computer open or supervise every step. It is more like delegating a task to an external service than collaborating in real time. Google positioned it as an alternative to tools like Devin, but with more direct native GitHub integration.
Companies
Tools
—
Tags
Sources