Multi-agent setup

Run several agents safely by giving each task its own lane and review path.

The safest multi-agent setup is simple: one lane per task, one clear goal per agent, and a review pass before merge.

Recommended pattern

1. Split the work

   Turn a large goal into small lane-sized tasks.

2. Create one lane per task

   Name lanes after the outcome, not the agent.

3. Start one chat per lane

   Keep each prompt scoped to that lane’s goal.

4. Watch Work and Graph

   Use Work for live sessions and Graph for branch relationships.

5. Review independently

   Inspect each diff, run tests, and create PRs separately.

Good task splits

Goal	Better split
”Improve onboarding”	Copy fixes, provider validation, first-run UI, tests.
”Fix checkout bugs”	Repro failing case, patch backend, patch UI, add regression tests.
”Ship mobile sync polish”	Desktop status, iOS display, reconnect behavior, docs.

Keep agents from colliding

• Do not send two agents into the same lane for unrelated work.
• Avoid broad prompts that invite repo-wide edits.
• Use stacked lanes when one task depends on another.
• Merge parent work before rebasing children when possible.
• Use History to understand what each session changed.

Lanes

Isolate each task.

Workspace graph

See how parallel work relates.