AgentWrapper / agent-orchestrator

A meta-harness that turns chaotic parallel AI coding agents into an orchestrated fleet with isolated worktrees and automatic CI/review feedback loops

正常维护 Apache 2.0 Go Tracked
8.6k 1.3k 1 个月前
CINPM
claude-code codex-cli orchestration orchestrator skills agent-fleet agent-swarm git-worktrees multi-agent parallel-agents parallel-coding tmux

星标趋势

数据积累中,暂无足够数据生成趋势图

AI 分析

项目摘要

Agent Orchestrator is a meta-harness IDE that supervises parallel AI coding agents in isolated git worktrees, providing automatic feedback loops for CI failures, review comments, and merge conflicts. It supports multiple agent CLIs including Claude Code, Codex, Cursor, and Kimi Code through a unified orchestration layer.

为什么值得关注

It solves the real coordination problem of running multiple AI coding agents in parallel — managing branches, terminals, CI feedback, and merge conflicts across agent sessions — turning chaotic parallel agent work into a managed workflow with 8,486 stars and very active development.

优势

  • Isolated git worktree per agent session prevents branch conflicts
  • Automatic feedback loops route CI failures, review comments, and merge conflicts back to the correct agent
  • Unified supervisor for multiple agent CLIs (Claude Code, Codex, Cursor, Kimi Code, opencode)
  • Very active development with 55 releases in 6 months and 42 contributors

局限性

  • High open issue count (404) suggests ongoing stability or feature completeness concerns
  • No Docker support and no examples provided, which may slow onboarding
  • Relatively new project (created Feb 2026) with limited production track record

使用场景

  • Orchestrating parallel AI coding agents on large features across a development team
  • Automated CI fix loops where agents self-correct based on test failures
  • Managing merge conflict resolution across multiple concurrent agent sessions
目标用户: Development teams and engineers who use multiple AI coding agents in parallel and need orchestration, visibility, and automated feedback loops
学习曲线:
分析模型:LongCat-2.0 | 分析时间:1 个月前