Documentation

Run a team of coding agents.
Stay in control.

Your agents can ship faster than you can keep track. bench turns your repo into mission control: every task is a card, every agent works in its own isolated worktree, every change waits as a PR for your review. You see everything. Nothing merges without you.

Install into a repo — running in 2 minutes bench on GitHub ↗

no account, no service, no database. it is python 3 on your own machine.

~/your-repo first run · abridged

$install_block

No .task-manager/manager/local/.env yet — a few questions and bench writes one.
Enter takes the default in [brackets]; Ctrl-D skips the rest.

  solo or team? [solo]:
  which agent adapter? [claude]:
  what command runs this project's tests? [python3 -m unittest]:

Wrote .task-manager/manager/local/.env — every other setting is in there, commented; edit it any time.

  …

Task board for ~/your-repo/.task-manager/tasks
  http://127.0.0.1:26071/
  Ctrl-C to stop

You've got agents. Now you've got a coordination problem.

“Which tab was that in?” Every agent's task, branch and status on one board — not scattered across chat sessions that forget everything.

“They keep stepping on each other.” Each agent works in its own git worktree. Parallel work, zero collisions.

“Did something just merge?” Never. bench opens PRs and puts CI on the card — the merge button is yours alone.

Run a team of coding agents.Stay in control.

You've got agents. Now you've got a coordination problem.

Run a team of coding agents.
Stay in control.