Skip to content
EN
English 简体中文 soon 日本語 soon

Jules

Google's async coding agent

Visit official site

What Jules is

Jules is Google's autonomous AI coding agent. You hand it a task, it pulls the code into an isolated cloud environment, reads the project context, makes a plan and produces reviewable code changes you can pull back into your process.

It handles multi-step work such as fixing bugs, adding tests, updating dependencies and version upgrades, refreshing documentation and implementing small features.

What you can do with it

  • Delegate routine engineering chores
  • Get dependency upgrades done as a reviewable change
  • Add missing tests
  • Keep documentation current

Who it is for

  • Teams maintaining large repositories
  • Developers wanting chores handled asynchronously
  • Teams using GitHub review

What to watch out for

  • Security logic, core architecture and complex business rules still need human control
  • Clear task boundaries produce better output; vague requests produce vague changes
  • Repository access must be scoped
  • Treat it as supervised automation, not an unsupervised system

Pros & cons

✓ What we like

  • Handles chores without blocking you
  • Produces reviewable changes
  • Fits GitHub review flow

! What to watch out for

  • Human control still required
  • Vague tasks give weak results
  • Access scope to set

FAQ

What does it do?

Runs engineering tasks such as bug fixes, tests, dependency updates and documentation in an isolated environment.

Is it only suggestions?

No. It produces actual code changes you review before merging.

What suits it?

Clearly bounded, fairly standard tasks.

Last reviewed: 2026-09-17

More AI coding tools tools

View all →

How we review