Maxim MCP Server

A Model Context Protocol (MCP) server implementation for Maxim AI that enables AI models to interact with Maxim's capabilities through a standardized interface.

🚀 Quick Start

Installation

You can install and run the Maxim MCP server using npx:

npx -y @maximai/mcp-server@latest

Environment Variables

The following environment variables are required to use the Maxim MCP server:

• MAXIM_API_KEY: Your Maxim API key for authentication
• MAXIM_BASE_URL: (Optional) The base URL for the Maxim API. Defaults to the production API URL.
• MAXIM_WORKSPACE_ID: (Optional) Your Maxim workspace ID if you want to scope the MCP server to a specific workspace.
• MAXIM_ENABLED_FEATURES: (Optional) Comma-separated list of features you want to enable. If not provided, all supported features will be enabled. Only the features that are supported by the server will be enabled, invalid feature names will be ignored.

Available features:

• dataset - Dataset management capabilities
• evaluator - Evaluator management and execution
• log_repository - Log repository management and insights

Configuration for Cursor

To use the Maxim MCP server with Cursor, add the following configuration to your Cursor settings:

{
    "mcpServers": {
        "maxim": {
            "command": "npx",
            "args": ["-y", "@maximai/mcp-server@latest"],
            "env": {
                "MAXIM_API_KEY": "your-api-key-here"
                // optionally you can also put MAXIM_WORKSPACE_ID to scope MCP server to a workspace
                // and MAXIM_ENABLED_FEATURES to only enable specific features.
            }
        }
    }
}

🔧 Features

The Maxim MCP server provides the following capabilities:

• Log Repositories
  • Create, view, update and delete log repositories
  • View and analyze logs of sessions and traces
  • Search through logs with advanced filtering options
  • Filter logs by timestamps, evaluator metrics, and trace properties
  • Organize log repositories in folders
• Datasets
  • Create, view, update and delete datasets
  • Manage dataset structure and content
  • Organize datasets in folders
• Evaluators
  • View evaluators
  • Execute evaluators with custom variables

📝 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.