mcpField
M
monnet

monnet

Local
monnet.aiProductivity0Updated 6w ago

About monnet

@monnet/mcp connects MCP-compatible clients to a Monnet workspace, exposing motions, plans, and approvals. It includes tools for reading and updating workspace items, with confirmation required before messages are sent under the user's identity.

Key features

  1. 01List workspaces and inbox items
  2. 02Read and update motions
  3. 03Create draft motions from prompts
  4. 04Approve or reject plan steps
  5. 05Confirm messages before sending

Use cases

  1. 01Review motions from an AI client
  2. 02Draft or revise a workspace motion
  3. 03Approve or reject a plan step
  4. 04Post a comment after confirmation

MCP support

This MCP server exposes Monnet workspace operations to compatible clients such as Claude Desktop and Cursor. Send actions use MCP elicitation for confirmation and abort if the client does not support it.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"ai-monnet": {
"command": "npx",
"args": [
"-y",
"@monnet/mcp"
],
"env": {
"MONNET_API_KEY": "<MONNET_API_KEY>"
}
}
}
}
Run commandnpx -y @monnet/mcp
Environment variables to set
MONNET_API_KEYSecretYour Monnet API key (mnk_...). Generate one at https://app.monnet.ai/settings (API Keys tab).

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@monnet/mcp0.9.14

Tools

33

Saved tool list · updated 2w ago

This tool takes no parameters.

Details

From the project README

@monnet/mcp

Model Context Protocol (MCP) server for Monnet.

Exposes motions, plans, and approvals to MCP-compatible clients (Claude Desktop, Cursor, etc.) so you can work with your Monnet workspace from your terminal without leaving your AI assistant.

Install

Add this to your MCP client config (e.g. ~/Library/Application Support/Claude/claude_desktop_config.json for Claude Desktop):

{
  "mcpServers": {
    "monnet": {
      "command": "npx",
      "args": ["-y", "@monnet/mcp"],
      "env": {
        "MONNET_API_KEY": "mnk_..."
      }
    }
  }
}

Generate an API key at https://app.monnet.ai/settings (API Keys tab).

Environment variables

VariableRequiredDefaultPurpose
MONNET_API_KEYyes—API key (mnk_*), sent as the X-API-Key header on every request
MONNET_API_URLnohttps://api.monnet.aiOverride for self-hosted or local dev backends

Tools

ToolPurpose
monnet_whoamiVerify the connection and list the API keys on your account
list_workspacesList the workspaces you belong to
get_inboxFetch your inbox
list_motionsList motions in a workspace
get_motionRead a motion — summary, body, plan, members, comments
create_motionCreate a new draft motion from a free-form prompt
update_motionUpdate a motion's summary, body, priority, or plan
commentPost a comment on a motion (approval-gated, see below)
approveApprove a plan step
rejectReject a plan step (with an optional reason)
ask_monnetAsk Monnet a question on a motion

Approval gate on send actions

Tools that send a message under your identity (e.g. comment) never dispatch automatically. Before the message leaves the MCP server, your client shows a confirmation dialog (MCP elicitation) with a preview; nothing is sent unless you explicitly approve. If your MCP client does not support elicitation, the send is aborted — the gate fails closed.

Local development

npm install
npm run build

Point your MCP client at the local build:

{
  "mcpServers": {
    "monnet": {
      "command": "node",
      "args": ["/absolute/path/to/monnet-mcp/dist/index.js"],
      "env": {
        "MONNET_API_KEY": "mnk_...",
        "MONNET_API_URL": "http://localhost:8000"
      }
    }
  }
}

Restart your MCP client after editing the config.

License

MIT

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.