Mission control for every repo you're shipping.
You don't work in one repo — you've got eight, each with failing checks, open PRs, half-finished agent sessions. Commandem is the desktop command center that shows what needs you across all of them, and dispatches the work back.
go install github.com/Scarcrux/commandem/cmd/commandem@latest
The interface
Not another coding agent. The window that runs them all.
Every tool is great inside one project and blind across them. Commandem lives above all of them — one surface for status, review, planning, and dispatch.
Plan
One roadmap across every project
A cross-project kanban board. Drag a card to restatus and it writes straight back to that project's roadmap, comment-preserving.
- Later / Next / Now / Done, every repo in one view
- Task checklists, due dates, and GitHub-issue links
- Author locally, sync up — never the system of record you can't leave
Review
Review agent work like a teammate
Every PR awaiting you, across repos, in one inbox. Open the diff, get an AI review, and post it back to GitHub as a real review — approve, comment, or request changes.
- PRs from all your projects, not one repo at a time
- AI review that becomes a genuine peer-review participant
- Secret-redacted diffs, staging, and inline commentary
Context
Every agent session, one map
claude, codex, and gemini sessions across all your repos, drawn as a single graph. Jump into any project's chat, code, activity, and roadmap from one window.
- Agent-agnostic — coordinate them, don't marry one vendor
- Per-project chat with tools, diff review, and recipes built in
- Cross-project memory: recall how you solved it in another repo
Insight
Know where the work actually went
Commits, finished agent runs, check transitions, and deploys — an activity feed and rollups across projects, plus a weekly digest. No timesheet.
- A cross-project activity feed, newest first
- Per-project throughput at a glance
- Real Claude usage in the menu bar, from your subscription
Why it's different
Four things a single-repo tool can't do
Because they live inside one project — and this lives above all of them.
What needs you, everywhere
One attention-bucketed view across every repo: failing checks, PRs to review, agents awaiting approval. Answers "what should I do right now?", not "what's happening."
Route the best agent per project
claude, codex, gemini, copilot, aider — per project or per task. Commandem coordinates them; it isn't tied to any one vendor's agent.
Run to a goal without babysitting
Autopilot works in a throwaway git worktree (never your tree), refuses "done" until your tests pass, and is seeded with how you solved it in your other repos. Guardrails before autonomy.
Commandem never sees your code
A single Go binary — no accounts, no hosted backend, no telemetry. Your registry, chats, and memory live on your disk, and you talk to the model provider directly, with your own subscription or key. We honestly can't say "your code never leaves your machine" — frontier models run in their cloud, under their terms. What we can say: nothing is routed through us, and you keep the receipts. What actually leaves →
Under the hood
The same power, fully scriptable
The app owns no business logic — it's a shell over one local binary. Everything the window does is also a keyboard-first command, so it drops straight into your shell, pipes, and cron.
$ commandem status # "what needs me right now?" across every repo $ commandem open backend --agent codex # route the best agent per project # Run it unattended, safely: isolated git worktree, gated on tests, # seeded with how you fixed this in your other repos. $ commandem autopilot api "fix the retry race" --worktree --verify "go test ./..." $ commandem recall "flaky retry" # how did I solve this before? (any repo) $ commandem status --json | jq # --json on every verb; ~80 of them
The marketplace
One marketplace for your agents — recipes, plugins, and skills
Three kinds of capability in one place: recipes (prompt templates rendered against your real repo context), plugins (sandboxed executables), and skills (capability packages your agent loads natively). Open, self-hostable, public artifacts only.
Browse the marketplace
Search recipes, plugins, and skills together — filter by kind, category, and source; copy or install any one. Community submissions are moderated before they're installable.
Open the marketplace →Measured, not piled
Every skill can carry a measured lift — a real, per-model verdict (HELPS +Npt, or an honest INERT) from an A/B over its own eval, so the shelf ranks by what actually works instead of by vibes.
One binary. Every repo. On your machine.
Install it, point it at your projects folder, and see what needs you.
commandem scan && commandem status