rescale
MCP server from rescale-labs/rescale-mcp-server
claude mcp add --transport stdio rescale-labs-rescale-mcp-server python -m src.rescale_api_mcp_server.cli \ --env RESCALE_BASE_URL="Base URL for Rescale API (defaults to https://platform.rescale.com/api/v3)" \ --env RESCALE_API_TOKEN="Rescale API token for authentication (required for API access)"
How to use
This MCP server exposes Rescale's Platform API through Claude Desktop and other MCP clients, enabling you to manage jobs, files, and resources directly from conversational interfaces. The server runs a Python module that implements the MCP protocol and talks to Rescale using your API token. Once running, Claude and other MCP clients will discover a Rescale toolset that can list, create, monitor, and manage compute jobs, upload and download files, and browse available hardware and software options. The examples show how to start the server in HTTP-like style and interact with it via the provided tools or clients. To use Claude Desktop with this server, configure the given mcpServers entry with your token and base URL, then restart Claude so the Rescale toolset appears in your workspace. The available features cover job management, file operations, resource discovery, and detailed documentation for advanced usage.
How to install
Prerequisites:
- Python 3.11+
- Git
- pip (comes with Python)
Step-by-step installation:
- Install Python 3.11+ and Git on your system.
- Create and activate a virtual environment:
- macOS/Linux: python3 -m venv venv source venv/bin/activate
- Windows: python -m venv venv venv\Scripts\activate
- Install the MCP server directly from GitHub: pip install git+https://github.com/rescale-labs/rescale-mcp-server.git
- Prepare environment variables (see next section) and run the server:
- Example to run via module entry point (see usage in the README): python -m src.rescale_api_mcp_server.cli --token YOUR_TOKEN --base-url https://platform.rescale.com/api/v3 --transport http --port 8000
Optional: You can also run the server in a development context by cloning the repo and running local source tree commands, if you prefer to modify or extend the codebase during development.
Additional notes
Tips and notes:
- Ensure you have a valid Rescale API token. You can export RESCALE_API_TOKEN or pass it via the CLI as shown in examples.
- RESCALE_BASE_URL can be used to point to different Rescale environments; it defaults to https://platform.rescale.com/api/v3.
- The MCP server module to run in Python is typically invoked via the CLI entry point: -m src.rescale_api_mcp_server.cli.
- If you encounter environment-related issues, verify that the virtual environment is activated and that Python 3.11 is being used.
- The README indicates Claude Desktop integration; after configuring your token and path, restart Claude Desktop to see the new Rescale tools.
- For testing, unit tests do not require real API calls, while integration tests require a valid RESCALE_API_TOKEN.
Related MCP Servers
mcp-vegalite
MCP server from isaacwasserman/mcp-vegalite-server
github-chat
A Model Context Protocol (MCP) for analyzing and querying GitHub repositories using the GitHub Chat API.
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
pagerduty
PagerDuty's official local MCP (Model Context Protocol) server which provides tools to interact with your PagerDuty account directly from your MCP-enabled client.
futu-stock
mcp server for futuniuniu stock
mcp -boilerplate
Boilerplate using one of the 'better' ways to build MCP Servers. Written using FastMCP