Modal

Provides a Modal-based MCP server to deploy and run Python scripts as Modal applications via a controllable MCP interface.

python

4

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "flux159-mcp-server-modal": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/mcp-server-modal",
        "run",
        "modal-server"
      ]
    }
  }
}

You deploy and run an MCP Server that enables you to host and execute Python scripts on Modal, coordinating with a filesystem backend to manage files. This guide walks you through installing prerequisites, configuring the MCP connections, and starting both the Modal-based server and the filesystem backend from your environment.

How to use

You will connect an MCP client to the Modal-based server and then deploy Python scripts to Modal through the MCP workflow. After you provide a script, you’ll receive a link to the modal application you can run, test, and share. The filesystem MCP server runs alongside to manage directories and file operations, enabling you to work with local files as part of the deployment and execution flow.

How to install

Prerequisites you need before starting: you should have Python and Modal installed and configured on your system.

pip install modal
python3 -m modal setup

Additional sections

MCP connections for this setup include two local stdio endpoints that you run from your environment. These endpoints are configured to be managed by your MCP client application, allowing you to deploy and coordinate Python scripts within Modal and to interact with the filesystem backend.

Example configuration for your MCP client (two stdio servers): one to run the Modal-based server and another to run the filesystem backend. Use exact commands and arguments as shown.

Configuration details for MCP connections

The following two stdio servers are used for this setup. They run locally and are started through your MCP client as part of your workspace.

Notes on development utilities

During development you can inspect the MCP server using the inspector tool to verify the runtime environment and deployed modal server. This helps ensure your endpoints are correctly wired and the modal server starts as expected.

Available tools

deploy_modal_app

Deploys a Python script to Modal by generating a Modal application, providing a shareable link, and enabling execution within the Modal environment.

start_modal_server

Starts the local MCP server that exposes the Modal-based APIs to the MCP client, enabling deployment, sharing, and running of Modal apps.