OrchestratorInc/agent-orchestrator

Run and supervise teams of coding agents from planning to merge. Any harness (Claude code, codex, +25 more). Desktop, web, mobile, and cloud agents.

What it solves

Agent Orchestrator (AO) addresses the complexity of managing multiple AI coding agents working on a single project. While a single agent can handle a task, running several in parallel creates challenges in coordinating work, preventing branch collisions, and tracking progress across various terminals, branches, and pull requests.

How it works

AO is a local desktop workspace that acts as a supervisor for coding agents. It uses a local daemon to monitor agent activity and source-control state. For Git-backed work, it provides each worker agent with its own isolated branch and worktree.

It features two primary levels of operation:

  • The Project Orchestrator: A persistent planning agent that handles high-level strategy, breaks ambiguous goals into concrete plans, and delegates tasks to worker agents.
  • Workers: Individual execution units that handle specific tasks using a chosen agent and model, operating within their own isolated workspace.

Progress is tracked via a live Kanban board that derives status from session data, pull requests, and CI results, grouping workers into states like "Working", "Needs you", "In review", and "Ready to merge".

Who it’s for

Developers who use AI coding agents (such as Claude Code, Aider, Cursor, or Devin) and want a centralized system to plan, run, and supervise multiple agents working on a large codebase.

Highlights

  • Isolated Execution: Every Git-backed worker gets its own branch and worktree to prevent collisions.
  • Project-Aware Planning: A dedicated orchestrator agent for high-level technical strategy and delegation.
  • ** uma live Kanban board for operational visibility into agent activity and PR status.
  • Integrated Tooling: Includes an agent-controllable browser for previews and support for both structured Chat and native terminal UIs.
  • Broad Agent Support: Compatible with a wide array of coding agents including Claude Code, Aider, and GitHub Copilot.

Related

  • Project
    storytold/pdfcraft

    An open-source, Rust-based PDF workbench that reimplements Adobe Acrobat's core features for reading, editing, and organizing PDFs with a focus on privacy and agent-automation.

  • Project
    InsForge/InsForge

    An open-source backend platform for agentic coding that provides AI agents with database, auth, storage, and compute to build full-stack apps autonomously.

  • Project
    Portkey-AI/gateway

    An open-source AI gateway that provides a unified API to route requests to 1,600+ LLMs and multi-modal models, featuring automatic retries, fallbacks, and guardrails.

  • Project
    ccfos/nightingale

    Nightingale is an open‑source, self‑hosted alerting platform that aggregates alerts from multiple monitoring data sources (Prometheus, VictoriaMetrics, Elasticsearch, Loki, ClickHouse, PostgreSQL, etc.). It provides a UI/API for rule management, team‑based permissions, event processing pipelines, and native notification integrations (Slack, PagerDuty, Discord, Email, Telegram, Jira, …). A built‑in AI assistant can query monitoring data, explain alerts, generate or modify rules, and run remediation scripts via extensible “Skills”. Deployable via Docker Compose, Linux binaries, or Helm on Kubernetes, Nightingale is Apache‑2.0 licensed and backed by a community Slack and GitHub discussions.