Anaconda MCP Server

Installing Anaconda MCP Server and connecting to Claude Desktop

Install the Anaconda MCP Server package

Open Terminal.
Create a new conda environment and install Anaconda MCP Server:
```
conda create --name <ENV_NAME> anaconda-mcp
```
Activate the environment:
```
conda activate <ENV_NAME>
```

When Anaconda MCP Server starts, it will open a browser window to Anaconda’s login page. However, authentication is not required to use Anaconda MCP Server.

Connect to Claude Desktop

You can connect to Claude Desktop automatically using the built-in setup command or by editing the configuration file manually.

Connect automatically (recommended)
Manual configuration

Activate the conda environment where you installed Anaconda MCP Server and run:

anaconda-mcp claude-desktop setup-config

If your Claude Desktop configuration file is in a non-standard location, you can specify the path to the configuration file:

anaconda-mcp claude-desktop setup-config -c <PATH_TO_CONFIG>

If you prefer to edit the Claude Desktop configuration file manually:

In Terminal, activate the conda environment where you installed Anaconda MCP Server.
Open Claude Desktop.
Navigate to Settings.
Select the Developer tab.
In the Local MCP Servers section, select Edit Config.

Open the claude_desktop_config file and add the following JSON configuration.

Finding your Python interpreter path

Open Terminal and run which python in the environment where you installed the anaconda-mcp package.

    {
        "mcpServers": {
            "anaconda-mcp": {
                "command": "/path/to/python",
                "args": ["-m", "anaconda_mcp", "serve", "--delay", "5"],
                "env": {
                    "ANACONDA_MCP_PYTHON_EXECUTABLE": "/path/to/python"
                }
            }
        }
    }

Configuration options

The env section of the Claude Desktop configuration supports the following options:By default, Anaconda MCP Server uses the mcp_compose.toml file in the directory where anaconda-mcp was installed. If your configuration file is in a non-standard location or has a custom name, use the --config flag and add the path to the configuration file to the args list:

    {
        "mcpServers": {
            "anaconda-mcp": {
                "command": "/path/to/python",
                "args": ["-m", "anaconda_mcp", "serve", "--delay", "5", "--config", "/path/to/mcp_compose.toml"],
                "env": {
                    "ANACONDA_MCP_PYTHON_EXECUTABLE": "/path/to/python",
                }
            }
        }
    }

Restart Claude Desktop

Quit and restart Claude Desktop for the changes to take effect.

Verify the configuration

Ask Claude something like “List my conda environments” to verify that Anaconda MCP Server is connected and responding to requests.

You can add “In conda” or “Using the conda server” to your prompts to ensure requests are routed to the conda MCP server.

Additional resources

Anaconda MCP Server is a full-featured MCP server that can be used with any MCP-enabled AI coding assistant; the Claude Desktop integration covered on this page is one way to use it. For full CLI reference documentation, including commands for starting and managing Anaconda MCP Server, composing MCP servers, and configuring client integrations, see the Anaconda MCP Server CLI reference.

Anaconda Tools

Anaconda MCP Server

Prerequisites

Capabilities

Installing Anaconda MCP Server and connecting to Claude Desktop

Additional resources

Troubleshooting

Anaconda Tools

​Prerequisites

​Capabilities

​Installing Anaconda MCP Server and connecting to Claude Desktop

​Additional resources

​Troubleshooting

Prerequisites

Capabilities

Installing Anaconda MCP Server and connecting to Claude Desktop

Additional resources

Troubleshooting