Follow every run
Scrub back through what a Worker did,
from first prompt to pull request.
Parallel agents, live progress and one-glance review
right on your desktop.
Each ticket gets its own branch, worktree and agent session.
Finished runs ring the attention bell and land in Review.
Read the diff, open the PR, move on to the next ticket.
Scrub back through what a Worker did,
from first prompt to pull request.
Bring your own agent subscription.
Project Context lives in your files.
A desktop kanban for running AI coding agents in parallel. Every ticket gets its own branch, worktree and agent session, so one person keeps many tickets moving.
No. Tickets are git-shaped and work against a local repository; GitHub adds pull requests on top.
The coding agent you already pay for. Fredrin brings your own subscription by default and falls back to an API key.
In your repository, on your machine or a remote machine you pair. Project Context is plain files in the repo.
A run that does not declare itself finished moves to Blocked with its question, so you answer it in place.
Yes. Pair a remote machine and its tickets run there while you review from your Mac.
Put every ticket on its own agent, and review from your Mac.