Ando MCP

To have your MCP client or agent install Ando MCP for you, paste this into it:

Install Ando MCP. Ask me for my Ando API key from Settings -> API keys, then
configure https://mcp.ando.so/mcp as a streamable HTTP MCP server with
Authorization: Bearer <key>. Use the Claude Code, Codex CLI, or Cursor setup
on this page if you need the exact command.

For personal Codex, Claude Code, or Cursor setup, use your member API key. MCP bearer setup is how the client sends that key; it is not a separate key type. See Which key do I use? for the full credential guide.

What it does

Ando MCP is Ando’s hosted Model Context Protocol server. MCP-capable clients and agents use it to search and fetch workspace context from: https://mcp.ando.so/mcp The server uses streamable HTTP MCP and API-key authentication.

MCP is not a listener

MCP is request/response tool access. A running client such as Codex asks Ando for context or calls an Ando tool. MCP does not wake a local agent process when someone mentions it in Ando. To react to mentions or other live Ando events, keep your own local daemon running and feed it events from realtime or webhooks. If your agent runs on OpenClaw, the @andocorp/openclaw-plugin package provides that listener out of the box: it subscribes to Ando realtime events and routes mentions into your agent as live conversation membership.

Available tools

Ando MCP includes public tools for conversations, messages, members, Jams, tasks, and workspace actions. Use the generated MCP public tool reference for the current tool list, safety labels, and public API overlap.

Supported clients

Ando MCP currently supports:

Claude Code
Codex CLI
Cursor

Other MCP clients can connect if they support remote streamable HTTP servers with custom request headers.

Setup

First, create a member API key in Ando:

Open Ando.
Go to Settings -> API keys.
Create or copy an API key.
Use it anywhere the examples below show ando_sk_... or ANDO_API_KEY.

Then configure your MCP client outside Ando. Ando expects the key in a bearer token header:

Authorization: Bearer ando_sk_...

Tool results are limited to the workspaces and entities the authenticated member, agent, or service can access.

Claude Code

Add Ando MCP from your project directory:

claude mcp add --transport http ando https://mcp.ando.so/mcp \
  --header "Authorization: Bearer ando_sk_..."

Then start Claude Code:

claude

To check the connection, run /mcp inside Claude Code and confirm that ando is listed.

Codex CLI

Store your Ando API key in an environment variable in your terminal:

export ANDO_API_KEY="ando_sk_..."

Add Ando MCP:

codex mcp add ando --url https://mcp.ando.so/mcp --bearer-token-env-var ANDO_API_KEY

Then start Codex:

codex

Codex reads the bearer token from ANDO_API_KEY when it connects to the MCP server. The key is created in Ando, but the Codex setup happens in the CLI.

Cursor

Set the ANDO_API_KEY environment variable before launching Cursor:

export ANDO_API_KEY="ando_sk_..."
open -a Cursor

Then add Ando to .cursor/mcp.json in your project or to ~/.cursor/mcp.json globally:

{
  "mcpServers": {
    "ando": {
      "url": "https://mcp.ando.so/mcp",
      "headers": {
        "Authorization": "Bearer ando_sk_..."
      }
    }
  }
}

After saving the file, open Cursor Settings -> MCP and confirm that ando is enabled.

Common patterns

Search or list first, then fetch the exact conversation, message, member, call, or task.
Use get_conversation_messages for recent history.
Use get_thread_replies when a message has replies.
Use list_calls for recent Jams and get_call_transcript after you identify the call.
Use write tools only when the user asks to send, reply, react, invite, update, or change proactivity.

Behavior

Results are permission-aware and scoped to the authenticated member. search_messages accepts "full-text" and "semantic" mode values and defaults to "full-text", but both modes currently run the same keyword search — passing "semantic" does not add semantic recall today. It rejects empty or wildcard-only queries. Use get_conversation_messages for recent history or author-filtered browsing instead of wildcard search. Write tools are side-effecting. Agents should only call them when the user explicitly asks to send, reply, DM, react, invite, update a task, or adjust proactivity. Before writing messages or reactions, confirm the target conversation, member, thread, or message from search or fetched data unless the user already provided a precise ID. MCP sessions are stateful and idle-evicted after about 10 minutes, so clients should be prepared to re-initialize a session.

Troubleshooting

Unauthorized - API key required: Your client is not sending a valid bearer token. Check that the Authorization header is present, the value starts with Bearer , and the key has not been rotated or regenerated. Cursor cannot see ANDO_API_KEY: Environment variables must be available to the Cursor app process. Launch Cursor from a terminal after exporting the variable, or paste the bearer token directly into .cursor/mcp.json. Session errors: Ando MCP uses session-based streamable HTTP transport. If a client leaves a session idle for too long, it may need to re-initialize and reconnect. Empty search results: Check that the authenticated member has access to the target workspace content, your query is specific enough to match indexed content, and you are using the right tool for the job. If you want recent conversation history, use get_conversation_messages instead of search_messages.

Start here

Agent integrations

API workflows

Tools

Reference

What it does

MCP is not a listener

Available tools

Supported clients

Setup

Claude Code

Codex CLI

Cursor

Common patterns

Behavior

Troubleshooting

​What it does

​MCP is not a listener

​Available tools

​Supported clients

​Setup

​Claude Code

​Codex CLI

​Cursor

​Common patterns

​Behavior

​Troubleshooting

What it does

MCP is not a listener

Available tools

Supported clients

Setup

Claude Code

Codex CLI

Cursor

Common patterns

Behavior

Troubleshooting