Concord MCP
Cross-harness communication and shared work-state for AI coding agents.
At a glance
Not tested yet- GitHub stars
- 1,462
- Weekly downloads
- 657
- Last commit
- 2026-11-20
- Licence
- MIT
- Transport
- stdio
- Auth
- none
npx -y @concord-ai/concord-mcp
What it does
Concord MCP is an open-source, local-first communication and coordination layer for AI coding agents. It lets agents in different harnesses send messages, detect overlapping work before editing, share decisions, and hand off tasks with evidence through one MCP server. It is for people running coding agents such as Claude Code, Codex, Cursor, Gemini CLI, Grok Build, Goose, GitHub Copilot CLI, and OpenCode. Concord is not another autonomous agent or orchestrator; it is the shared layer around your agents.

Tools it gives your agent
start_work: registers presence, claims or accepts one task, and reports scope overlaps before editing.inspect_work: reads workspace or task state, an agent inbox or outbox, or a durable prompt and reply thread.update_work: records task context or immediately prompts or replies to another promptable workspace agent.transfer_work: assigns, accepts, declines, releases, reassigns, offers handoffs, or reopens versioned work.finish_work: records evidence and optionally marks a task review-ready, complete, or closed.
Quick start
npm install -g @concord-ai/concord-mcp cd /path/to/your/repository concord setup
Restart your agent clients, then ask two of them to work in the same repository. On macOS or Linux without Node.js, the sources give a self-contained install instead:
curl -fsSL https://getconcord.ai/install.sh | sh
Add it to Claude Code from a terminal:
claude mcp add concord-mcp -- npx -y @concord-ai/concord-mcp
Or add this to .mcp.json in your project root (Claude Code) or .cursor/mcp.json (Cursor):
{
"mcpServers": {
"concord-mcp": {
"command": "npx",
"args": [
"-y",
"@concord-ai/concord-mcp"
]
}
}
}Environment variables
CONCORD_INSTALL_DIR: overrides where the standalone build is installed.CONCORD_BIN_DIR: overrides where theconcordandconcord-mcplaunchers are placed.CONCORD_VERSION: pins a release of the standalone build.
Example prompts
- Claim
src/app/page.tsxbefore you start editing, and tell me if another agent already owns it. - Send a prompt to the other agent in this repository asking which files it plans to take.
- Record the decisions and findings from this task, then mark it review-ready with the test evidence.
- Hand this task off to another agent and include the scope, tests, and risks.
What to watch out for
- Live delivery depends on the receiving harness and session state; hook-only integrations leave a durable pull message and state that limitation in the result.
- Delivery fails immediately when the named agent has no reachable endpoint; Concord does not silently reroute it.
- There is no universal
/concordslash command; commands are client-specific. - Agents on different machines require signing in to Concord Cloud with
concord cloud login. - The README lists no costs.
Not tested by tools4agents yet. Read the source repository before you install it.
Similar tools
Everything
The official test server. Exercises every part of the protocol, so it is the fastest way to check your setup works at all.
Filesystem
Gives an agent scoped read and write access to directories you name explicitly. The usual first MCP server people install.
Git
Read, search, and manipulate local Git repositories. Lets an agent inspect history and diffs without shelling out.
Sequential Thinking
Gives a model a scratchpad for multi step reasoning, letting it revise earlier steps instead of committing to a first answer.