Albert

Albert coordinates Claude Code agents for developers who want to hand off a coding goal that spans several tasks. It plans the work and continues through successive rounds until it reaches the goal or a requested review checkpoint.

• Parallel coding tasks: Albert splits goals into tasks, assigns Haiku, Sonnet or Opus, and gives independent agents separate Git worktrees so they can work concurrently.
• Independent checks: Separate agents rerun checks, review code and test user journeys and edge cases; research tasks also face a skeptic that challenges their conclusions.
• Pull request review: Albert can leave pull requests open for human review or merge them automatically after code review and QA pass, with deployments, migrations and other irreversible actions requiring explicit authorization.
• Local activity dashboard: The Albert Console shows agent status, exchanges between agents and run histories across projects, including task logs and token use.
• Optional chat: Albert Chat answers questions about runs, accepts changes to priorities or scope and starts new runs through your existing Claude Code login; steering messages can't weaken approval gates or reverse a reviewer's rejection.

The orchestration software and console run locally on Windows 10/11, macOS or Linux. Albert requires Node.js and an installed, signed-in Claude Code account. The console is read-only and accessible only from the local machine. It sends structured activity metadata to the browser, excluding prompt and response bodies; optional chat runs as a separate local service.