Skip to content

Jules

Jules is Google's asynchronous AI coding agent. Unlike interactive assistants that work while you watch, Jules operates in the background — assign it a GitHub issue, and it clones the repo, analyzes the problem, writes code, runs tests, and submits a pull request. You review the PR when it's ready. MCP tools give Jules access to external systems during these autonomous sessions.


Ju
Jules
Google · jules.google
TRANSPORT
Streamable HTTP ✓
PLATFORM
Web (GitHub-integrated)
MCP VIA
Agent Configuration

Background Agent for GitHub Issues

Jules' asynchronous model is its differentiator. You keep coding while Jules works on a separate task. Assign a bug fix or feature request, and check back later for a ready-to-merge PR with tests and documentation.

Powered by Google's Gemini models, Jules excels at understanding complex codebases and generating contextually appropriate changes.

Features:

  • Asynchronous — works in the background while you do other tasks
  • GitHub-native — assign issues, review PRs
  • Self-testing — runs tests and verifies its own changes
  • Plan preview — review the approach before execution starts
  • Gemini-powered — Google's advanced models for code understanding
  • Multi-file — coherent changes across the codebase

Setup

1. Create a Token

In Vinkius Cloud, go to your server → Connection TokensCreate. Copy the URL.

2. Configure MCP

Add your Vinkius URL in Jules' MCP configuration.

3. Assign and Wait

Assign a GitHub issue to Jules. Review the pull request when it's ready.


FAQ

Does Jules work while I code? Yes. Jules runs asynchronously. You continue your own work while Jules handles the assigned task.

Can I review Jules' approach before it starts? Yes. Plan preview shows the proposed strategy before execution.

How do I assign issues to Jules? Through GitHub issue assignment or the Jules web interface.

Is Jules free? Free tier available through Google.