mcpField
A
amtb-mcp

amtb-mcp

Local
github.com/l2yaoOther0Updated 27w ago
Repository

About amtb-mcp

amtb-mcp is a Python server for interacting with AMTB resources and tools through MCP, for users who want to call its functions from clients such as Claude Desktop. It runs locally over stdio and includes menu-fetching and document-conversion tools.

Key features

  1. 01Exposes MCP tools locally over stdio
  2. 02Fetches AMTB menus to files
  3. 03Provides Python tool wrappers
  4. 04Includes document conversion tools

Use cases

  1. 01Configure a local AMTB server in Claude Desktop
  2. 02Call tool wrappers directly during development
  3. 03Fetch menus and generate category files
  4. 04Convert .doc files with document tools

MCP support

The server exposes implemented tools over stdio for local use by MCP clients such as Claude Desktop.

#amtb#mcp#server#resources#tools

Configuration

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

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

amtb-mcp

MCP server for interacting with AMTB resources / tools

Use in Claude Desktop

To use it with Claude Desktop, add to your claude_desktop_config.json:

{
  "mcpServers": {
    "amtb": {
      "command": "python",
      "args": ["{your_path}/amtb-mcp/run_mcp.py"]
    }
  }
}

Quickstart

Requirements

  • Python 3.13+ (recommended)
  • pip
  • Optional: LibreOffice (soffice) or pywin32 for .doc → .docx conversion when using doc conversion tools.

Setup (Windows)

  1. python -m venv .venv
  2. ..venv\Scripts\Activate.ps1 # PowerShell or ..venv\Scripts\activate.bat # cmd

Install

  • pip install -e .

Run locally

  • Start the MCP server (stdio transport):

    python run_mcp.py

    This exposes the implemented tools on stdio for local testing or consumption by an MCP client (e.g., Claude Desktop).

  • Alternatively you can run the main module directly:

    python main.py

Testing

  • Run unit tests:

    python -m unittest discover -v

    or, if you prefer pytest:

    pytest

Using tools directly (for development)

  • Call wrappers from Python for quick checks, e.g.:

    python -c "from fetch_tool import fetch_amtb_menu; print(fetch_amtb_menu(write_files=False))"

    python -c "from main import get_amtb_categories; print(get_amtb_categories())"

Notes

  • fetch_amtb_menu writes menu.json and per-category files under out_dir (default: amtb). Use write_files=False for a dry run.
  • Download and conversion tools are opt-in and may require additional system packages and time.
  • For development, place new tool modules under amtb-mcp/ and register them in main.py (or import modules that call mcp.tool() on module import).

Contributing

  • Add a tool module, include tests, and submit a PR. Keep operations idempotent and return JSON-serializable Pydantic models.

License

  • See LICENSE in repository root.

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.