Three terminal recordings, from a cold install to five crews running in parallel. Everything runs locally; you bring your Anthropic key.
Install Maestro (the free Community tier needs no licence key), scaffold a workspace, create a voyage and an order, then dispatch your first crews. This is a real session, recorded from a clean install.
mso dispatch --max-crews N launches specialised crews against your active voyage, up to five at once (the Community tier runs two). Here, three crews pick up three orders together. mso status shows live progress: crew names, roles, model tier, turn count, and token spend.
Watch a real order progress through its role sequence, with the lifecycle log showing each hand-off. Each crew hands off to the next only when its acceptance criteria pass. Add REL to the sequence and the Releaser squad triages CI and the review threads, then signs off with a READY or NOT READY recommendation. Merging stays a human decision: Maestro never merges for you.
Everything above runs from your shell. The same fleet also reports into Maestro Hub, the web dashboard where voyages are tracked, plans are approved and token spend is visible across every project. A walkthrough of it is being recorded.
The demos show the shape of a run. The reference docs have the detail: every flag, every config key, and the SCM and CI providers Maestro can drive.
Install from PyPI, activate a licence, scaffold a workspace with mso setup, and bring up the Hub.
Read the install guide →
Every mso command grouped by task: dispatch, status, voyages, orders, usage, security, doctor.
Browse the commands →
Point Maestro at your stack. GitHub or Bitbucket for source, GitHub Actions or Jenkins for CI, configured per workspace.
Configure providers →
Stuck on something specific? The troubleshooting and FAQ page collects the problems operators actually hit, and how to get out of each.
Public wheel, no GitHub credentials. Dispatching squads inside 90 seconds.