Atla MCP Server

An MCP server implementation providing a standardized interface for LLMs to interact with the Atla API for state-of-the-art LLMJ evaluation.

Learn more about Atla here. Learn more about the Model Context Protocol here.

Available Tools

evaluate_llm_response: Evaluate an LLM's response to a prompt using a given evaluation criteria. This function uses an Atla evaluation model under the hood to return a dictionary containing a score for the model's response and a textual critique containing feedback on the model's response.
evaluate_llm_response_on_multiple_criteria: Evaluate an LLM's response to a prompt across multiple evaluation criteria. This function uses an Atla evaluation model under the hood to return a list of dictionaries, each containing an evaluation score and critique for a given criteria.

Usage

To use the MCP server, you will need an Atla API key. You can find your existing API key here or create a new one here.

Installation

We recommend using uv to manage the Python environment. See here for installation instructions.

Clone the repository:

git clone https://github.com/atla-ai/atla-mcp-server.git
cd atla-mcp-server

Create and activate a virtual environment:

uv venv
source .venv/bin/activate

Install dependencies depending on your needs:

# Basic installation
uv pip install -e .

# Installation with development tools (recommended)
uv pip install -e ".[dev]"
pre-commit install

Add your ATLA_API_KEY to your environment:

export ATLA_API_KEY=<your-atla-api-key>

Connecting to the Server

Once you have installed the server, you can connect to it using any MCP client.

Here, we provide specific instructions for connection to some common MCP clients.

In what follows:
If you are having issues with uv, you might need to pass in the full path to the uv executable. You can find it by running which uv in your terminal.
path/to/atla-mcp-server is the path to the atla-mcp-server directory, which is the path to the repository you cloned in step 1.

Having issues or need help connecting to another client? Feel free to open an issue or contact us!

OpenAI Agents SDK

For more details on using the OpenAI Agents SDK with MCP servers, refer to the official documentation.

Install the OpenAI Agents SDK:

pip install openai-agents

Use the OpenAI Agents SDK to connect to the server:

import os

from agents import Agent
from agents.mcp import MCPServerStdio

async with MCPServerStdio(
        params={
            "command": "uv",
            "args": ["run", "--directory", "/path/to/atla-mcp-server", "atla-mcp-server"],
            "env": {"ATLA_API_KEY": os.environ.get("ATLA_API_KEY")}
        }
    ) as atla_mcp_server:
    ...

Claude Desktop

For more details on configuring MCP servers in Claude Desktop, refer to the official MCP quickstart guide.

Add the following to your claude_desktop_config.json file:

{
  "mcpServers": {
    "atla-mcp-server": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/atla-mcp-server",
        "run",
        "atla-mcp-server"
      ],
      "env": {
        "ATLA_API_KEY": "<your-atla-api-key>"
      }
    }
  }
}

Restart Claude Desktop to apply the changes.

You should now see options from atla-mcp-server in the list of available MCP tools.

Cursor

For more details on configuring MCP servers in Cursor, refer to the official documentation.

Add the following to your .cursor/mcp.json file:

{
  "mcpServers": {
    "atla-mcp-server": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/atla-mcp-server",
        "run",
        "atla-mcp-server"
      ],
      "env": {
        "ATLA_API_KEY": "<your-atla-api-key>"
      }
    }
  }
}

You should now see atla-mcp-server in the list of available MCP servers.

Running the Server

If you are using an MCP client, you will generally not need to run the server locally.

Running the server locally can be useful for development and debugging. After installation, you can run the server in several ways:

Using uv run (recommended):

cd path/to/atla-mcp-server
uv run atla-mcp-server

Using Python directly:

cd path/to/atla-mcp-server
python -m atla_mcp_server

With the MCP Inspector:

cd path/to/atla-mcp-server
uv run mcp dev src/atla_mcp_server/debug.py

All methods will start the MCP server with stdio transport, ready to accept connections from MCP clients. The MCP Inspector will provide a web interface for testing and debugging the MCP server.

Contributing

Contributions are welcome! Please see the CONTRIBUTING.md file for details.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Integrations