Skip to content
MCP server

Lanes

Shared context and file queues so coding agents on one team stop overwriting each other

Runs locallyStatus and history
Pricing
Open source
Platforms
self-hosted, CLI, vscode
Runs
Locally
Version
0.3.0
Listing updated
3 days ago

About Lanes

Lanes is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.

Lanes starts from a narrow problem: several teammates each run a coding agent in the same repository, and nobody sees what the other agents are doing until a merge goes wrong. Before each edit an agent takes a short lease on the file. If a teammate's agent holds it, yours waits in a fair queue, then picks up the latest content and goes ahead. Lanes never refuses an edit; if the server is down or the wait runs out, the edit proceeds and is committed.

The optional team context goes further than a lease can. Each prompt starts with a short brief of who is working on what, and a one-line heads-up arrives before an agent reads or edits a file a teammate is changing, including that teammate's task. It is off until you turn it on per repository with lanes team on and commit .lanes/team.json. Requests to a teammate's agent reach it only after that person approves, and the MCP server has no approve tool.

It is self-hosted: one teammate runs the server on a laptop or VM. The server sees paths, names, statuses and task titles, never file contents, and sends no telemetry. The README reports 389 real Claude Code sessions and 108 benchmark runs, with Claude Code run live and the other 18 agent integrations built from their docs but not yet tested against the real agents.

Install

Run in a terminal

claude mcp add lanes -- npx -y https://github.com/mfmatozza/lanes/releases/download/v0.3.0/lanes-0.3.0.mcpb

Add to claude_desktop_config.json

{
  "mcpServers": {
    "lanes": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/mfmatozza/lanes/releases/download/v0.3.0/lanes-0.3.0.mcpb"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "lanes": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/mfmatozza/lanes/releases/download/v0.3.0/lanes-0.3.0.mcpb"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "lanes": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/mfmatozza/lanes/releases/download/v0.3.0/lanes-0.3.0.mcpb"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "lanes": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/mfmatozza/lanes/releases/download/v0.3.0/lanes-0.3.0.mcpb"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "lanes": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/mfmatozza/lanes/releases/download/v0.3.0/lanes-0.3.0.mcpb"
      ]
    }
  }
}

What it does

File leases and queue
An agent leases a file before editing; a second agent waits in a fair queue and then gets the newer version.
Team context brief
Each prompt carries a short note on who is working on what, and recent team-chat mentions.
Heads-up before edits
A one-line warning arrives before an agent reads or edits a file a teammate is changing.
Approved agent requests
Ask a teammate's agent to explain or do something; it arrives only after a person approves.
Never blocks edits
If coordination fails, the edit goes ahead and is committed.
Several views
See the team in the CLI, a browser app, or an optional VS Code sidebar.

Good for

  • Run several teammates' coding agents in one repository without overwritten changes
  • Warn an agent that a function it is calling is being renamed
  • Show a team who is working on which files right now
  • Pass a request from your agent to a teammate's agent with human approval

Pricing

Open source

Free and open source

Prices change. Check trylanes.app before you buy.

Lanes FAQ

Does Lanes ever block an edit?

No. If the server is down or a wait runs out, the edit goes ahead uncoordinated and is committed.

What data leaves my machine?

The server sees paths, names, statuses and task titles. File contents never leave your machine, and a task title is cut to 120 characters with secrets redacted.

Which agents work with it?

Claude Code is run live. Eleven more agents work through their own hooks or plugins and seven more through MCP only, but those 18 are built from each agent's docs and have not been run against the real agents yet.

Where does the server run?

It is self-hosted with no account; one teammate runs it on a laptop or a VM.

Is Lanes free?

Yes. Lanes is open source, so you can use and self-host it for free.

How do I add Lanes to Claude Code?

Run this command in your terminal: claude mcp add lanes -- npx -y https://github.com/mfmatozza/lanes/releases/download/v0.3.0/lanes-0.3.0.mcpb

What are the best alternatives to Lanes?

The closest alternatives to Lanes are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Lanes

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools