Albert

Albert coordinates Claude Code agents for developers who want to hand off a coding goal across several tasks. Its harness and dashboard run locally on Windows, macOS or Linux, and it requires an installed, signed-in Claude Code account. It assigns work to Haiku, Sonnet or Opus according to the task and continues until the goal is complete or it reaches a requested review checkpoint.

Independent tasks run in parallel in separate Git worktrees. Separate agents check the results: verification reruns checks from a clean tree, code review examines changes, and QA tests user journeys and edge cases. Research tasks also face a skeptic agent that challenges their claims. Albert can leave pull requests open for human review or merge them after review and QA pass. Deployments, migrations and other irreversible actions require explicit authorization.

The local browser console shows agent activity across projects, communication between agents, and run histories with task details, logs and token use. It's read-only. The console receives structured activity metadata without bringing prompt or response bodies into the browser. Optional Albert Chat lets you ask about run status, redirect ongoing work or launch another run. It reuses your Claude Code login without storing API keys; steering messages can't weaken approval gates or override a reviewer's rejection.

Albert is source-available under the PolyForm Noncommercial License 1.0.0, which permits noncommercial use, modification and sharing but restricts use for commercial advantage.