project-brain-mcp

Coordinate multiple AI agents across your codebase

> While Cursor handles one file and Claude Code helps with another, Project Brain tracks what each agent worked on, prevents conflicts, and makes it easy to hand off context between tools.

[](https://opensource.org/licenses/MIT) [](https://nodejs.org) [](https://modelcontextprotocol.io)

---

When you use multiple AI coding agents (Claude Desktop, Cursor, GitHub Copilot), it's hard to track:

• ❌ What each agent is working on
• ❌ Which conversations led to which code changes
• ❌ Important decisions made during development
• ❌ How to hand off context between agents

Project Brain solves this by:

• ✅ Task coordination - Assign work to specific agents
• ✅ Conversation search - Find past discussions instantly 🆕
• ✅ Decision tracking - Record architectural choices
• ✅ Context handoff - Export/import complete project state 🆕
• ✅ 100% local - Works offline, no cloud dependencies
• ✅ No vendor lock-in - Full import/export to JSON

Free Tier:

• ✅ Up to 100 tasks per project
• ✅ Full-text search across tasks, conversations, decisions
• ✅ Complete import/export for portability
• ✅ Works with any MCP-compatible AI tool

Upgrade for Teams:

• 🚀 Unlimited tasks and projects
• 🚀 Multi-agent coordination with task locking
• 🚀 Real-time collaboration dashboard
• 🚀 AI code review and analysis

---

Step 1: Configure your AI tool

Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, %APPDATA%\Claude\claude_desktop_config.json on Windows):

{
  "mcpServers": {
    "project-brain": {
      "command": "npx",
      "a