mai-tai MCP Server

MCP (Model Context Protocol) server for mai-tai, enabling AI coding agents to communicate with humans through the mai-tai platform.

Quick Start

1. Install with uv (recommended)

# One-time install (adds to PATH)
uv tool install mai-tai-mcp

# Or run directly without installing
uvx mai-tai-mcp

2. Alternative: Install with pipx

pipx install mai-tai-mcp

3. Development install

cd mcp-server
pip install -e .

Configuration for Claude Desktop

Add this to ~/.config/claude/claude_desktop_config.json (Mac/Linux) or the equivalent on Windows:

{
  "mcpServers": {
    "mai-tai": {
      "command": "uvx",
      "args": ["mai-tai-mcp"],
      "env": {
        "MAI_TAI_API_URL": "https://api.mai-tai.dev",
        "MAI_TAI_API_KEY": "mt_your_api_key_here",
        "MAI_TAI_PROJECT_ID": "your-project-id"
      }
    }
  }
}

Then restart Claude Desktop.

Configuration for Augment

Add to your Augment MCP settings:

{
  "mcpServers": {
    "mai-tai": {
      "command": "uvx",
      "args": ["mai-tai-mcp"],
      "env": {
        "MAI_TAI_API_URL": "https://api.mai-tai.dev",
        "MAI_TAI_API_KEY": "mt_your_api_key_here",
        "MAI_TAI_PROJECT_ID": "your-project-id"
      }
    }
  }
}

Environment Variables

Variable	Description	Example
`MAI_TAI_API_URL`	mai-tai backend URL	`https://api.mai-tai.dev`
`MAI_TAI_API_KEY`	API key from mai-tai project	`mt_xxx...`
`MAI_TAI_PROJECT_ID`	Your project ID	`df944ea7-...`

All three variables are required. The server will fail fast with a clear error message if any are missing.

Available Tools

`ask_human`

Primary tool for escalating questions to humans. Use when you need:

Information you don't have access to
A decision or approval
Clarification on requirements
Help with an error you can't resolve

`list_channels`

List all channels in the mai-tai project.

`get_channel`

Get details about a specific channel.

`send_message`

Send a message to a channel (non-blocking status updates).

`get_messages`

Get messages from a channel.

`get_project_info`

Get information about the connected mai-tai project.

Note: Agents cannot create channels. Channels must be created by humans via the mai-tai web UI.

Getting an API Key

Log in to mai-tai at https://mai-tai.dev
Go to your project settings
Create a new API key
Copy the key (it's only shown once)

Testing

Test with MCP Inspector:

MAI_TAI_API_URL="http://localhost:8000" \
MAI_TAI_API_KEY="your-key" \
MAI_TAI_PROJECT_ID="your-project-id" \
  npx @modelcontextprotocol/inspector mai-tai-mcp

Metadata

Release files for mai-tai-mcp 0.7.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for mai-tai-mcp 0.7.0
File	Size	Uploaded
mai_tai_mcp-0.7.0.tar.gz	36.9 kB	Aug 6, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for mai-tai-mcp 0.7.0
File	Interpreter	ABI	Platform	Reset
mai_tai_mcp-0.7.0-py3-none-any.whl 33.7 kB Aug 6, 2026	Python 3	none	any	Details

Total release size: 70.6 kB

Release files / mai_tai_mcp-0.7.0.tar.gz

Download URL	mai_tai_mcp-0.7.0.tar.gz
Size	36.9 kB
Tags	Source
SHA-256 checksum How to use checksums	`1d0217f3b6eb989f6f7ef6fb00f9208018ae8653a18f0f400b3ece5aa1b49f6a`
BLAKE2b-256 checksum How to use checksums	`fd2255e0640301a8a520718299522ab752f51bbbaa422f37cd6d60617b2ac82b`
Upload date	Aug 6, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/7.0.0 CPython/3.13.14`

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/jmcdice/mai-tai-dev
Publishing commit: github.com/jmcdice/mai-tai-dev/tree/57342d558b6733cea0f574fbaeb63aff93272a10

Workflow

Publishing configuration: github.com/jmcdice/mai-tai-dev/blob/57342d558b6733cea0f574fbaeb63aff93272a10/.github/workflows/publish-mcp.yml
Publishing logs: github.com/jmcdice/mai-tai-dev/actions/runs/31131915715/attempts/1

Artifact

Subject: mai_tai_mcp-0.7.0.tar.gz
SHA-256 checksum: 1d0217f3b6eb989f6f7ef6fb00f9208018ae8653a18f0f400b3ece5aa1b49f6a Verifying attestations

Release files / mai_tai_mcp-0.7.0-py3-none-any.whl

Download URL	mai_tai_mcp-0.7.0-py3-none-any.whl
Size	33.7 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`eabe2c2fd7aced5632bec74fed063a63a635d3781b8fc05529b8292845bdd510`
BLAKE2b-256 checksum How to use checksums	`45bbe3b03a6e8e3925b1d9f3b3e9bcb1d77b47363941c3be82af6cd3b7eb2689`
Upload date	Aug 6, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	Yes
Uploaded via	`twine/7.0.0 CPython/3.13.14`

Provenance

Identity

Publishing platform: GitHub Actions
Publishing repository: github.com/jmcdice/mai-tai-dev
Publishing commit: github.com/jmcdice/mai-tai-dev/tree/57342d558b6733cea0f574fbaeb63aff93272a10

Workflow

Publishing configuration: github.com/jmcdice/mai-tai-dev/blob/57342d558b6733cea0f574fbaeb63aff93272a10/.github/workflows/publish-mcp.yml
Publishing logs: github.com/jmcdice/mai-tai-dev/actions/runs/31131915715/attempts/1

Artifact

Subject: mai_tai_mcp-0.7.0-py3-none-any.whl
SHA-256 checksum: eabe2c2fd7aced5632bec74fed063a63a635d3781b8fc05529b8292845bdd510 Verifying attestations