Agent Collaboration MCP Server

Self-contained MCP server for multi-agent collaboration with automatic authentication delegation and complete tmux session management - no external dependencies required. The agent collaboration mcp mcp server wraps that behind the Model Context Protocol, so an assistant can use it through 14 defined tools rather than through you.

What it actually does

複数のAIエージェントが協調して作業できるようにする完全自己完結型MCPサーバーです。このツールをClaude Codeなどのエージェントに使わせることで、エージェントが他のエージェントを起動・制御し、チームとして複雑なタスクを実行できます。

Adding it to your client

agent-collaboration-mcp on npm is all you need. Most clients run it directly, so configuration is a few lines and a restart.

Its toolset

Everything the assistant can do here goes through one of these:

• target — tmuxターゲット形式（"session:window.pane"、例："multiagent:0.5"）
• agentType — claude（一般的なコード開発）または gemini（画像生成タスク）
• additionalArgs — 追加の引数（オプション）
• prompt — 実装指示（必須）
• workerCount — ワーカー数（デフォルト: 3）
• complexity — タスクの複雑度（simple, medium, complex）
• autoMerge — 完了後の自動マージ（デフォルト: false）
• tmux — tmuxセッション: tmux new-session -d -s multiagent
• agent_manager.sh — エージェント起動コマンドの定義
• pane_controller.sh — メッセージ送信方法の定義
• running_claude — ✅
• running_gemini — 💎

Caveats

• It runs with your machine's permissions. That is convenient and also the reason to think about what you point it at before you approve a tool call.
• With 14 tools registered it takes up a noticeable share of the context window; turn it off in projects that never touch Agent Collaboration MCP.
• MCP clients confirm each tool call by default. Leave that on until you have watched what the agent collaboration mcp mcp server does with a few real requests.

When to reach for it

This sits in the developer tooling group, where several servers overlap in what they claim to do but differ sharply once you actually set them up. Agent Collaboration MCP's toolset — target, agentType, additionalArgs and 11 more — is a fair guide to whether it matches your workflow. It is maintained by nishimoto265; worth a glance at recent repository activity before you build anything load-bearing on it.

We check each listing at SyncDev against the project's documentation before it goes live — if something here drifts out of date, it is a bug worth reporting.