MCP Servers

Connect AI assistants to Phoenix documentation and platform capabilities via the Model Context Protocol (MCP).

Phoenix provides two MCP servers for AI assistants:

• Phoenix Docs MCP - Search Phoenix documentation in real-time
• Phoenix MCP Server - Manage prompts, datasets, and experiments

---

Phoenix Docs MCP

The Phoenix Docs MCP server allows AI assistants to search and retrieve Phoenix documentation in real-time.

Server URL:

https://arizeai-433a7140.mintlify.app/mcp

Connect with Claude Code

# Project scope (current directory only)
claude mcp add --transport http phoenix-docs https://arizeai-433a7140.mintlify.app/mcp

To add the MCP server globally and access it in all projects, add the --scope user flag:

claude mcp add --transport http phoenix-docs --scope user https://arizeai-433a7140.mintlify.app/mcp

Connect with Cursor or VS Code

Add the following to your MCP settings configuration file:

{
    "mcpServers": {
        "phoenix-docs": {
            "url": "https://arizeai-433a7140.mintlify.app/mcp"
        }
    }
}

Connect with Claude Desktop

1. Open Claude Desktop
2. Go to Settings > Connectors
3. Add the MCP server URL: https://arizeai-433a7140.mintlify.app/mcp

---

Phoenix MCP Server

The Phoenix MCP Server (@arizeai/phoenix-mcp) connects AI assistants directly to your Phoenix instance for managing:

• Prompts Management: Create, list, update, and iterate on prompts
• Datasets: Explore datasets and synthesize new examples
• Experiments: Pull experiment results and visualize them with the help of an LLM

Connecting the Phoenix MCP Server

Connect via Cursor

Via Cursor Deeplink:

Add to Cursor

Manually:

From the Cursor Settings page, navigate to the MCP section, and click “Add new global MCP server”

Add the following code to your MCP config file:

{
  "mcpServers": {
    "phoenix": {
      "command": "npx",
      "args": [
        "-y",
        "@arizeai/phoenix-mcp@latest",
        "--baseUrl",
        "https://my-phoenix.com",
        "--apiKey",
        "your-api-key"
      ]
    }
  }

Replacing:

• https://my-phoenix.com with your Phoenix collector endpoint

• your-api-key with your Phoenix API key

After saving your config file, you should see the Phoenix server enabled:

You can access Phoenix prompts, experiments, and datasets through Cursor!

Connect via Claude Desktop

From the Claude Desktop settings window, navigate to the Developer Section, and click “Edit Config”

Open your config file and add the following code:

{
  "mcpServers": {
    "phoenix": {
      "command": "npx",
      "args": [
        "-y",
        "@arizeai/phoenix-mcp@latest",
        "--baseUrl",
        "https://my-phoenix.com",
        "--apiKey",
        "your-api-key"
      ]
    }
  }

Replacing:

• https://my-phoenix.com with your Phoenix collector endpoint

• your-api-key with your Phoenix API key

Save your file and relaunch Claude Desktop. You should now see your new tools ready for use in Claude!

Manually Connect

Add the following code to your MCP config file:

{
  "mcpServers": {
    "phoenix": {
      "command": "npx",
      "args": [
        "-y",
        "@arizeai/phoenix-mcp@latest",
        "--baseUrl",
        "https://my-phoenix.com",
        "--apiKey",
        "your-api-key"
      ]
    }
  }

Replacing:

• https://my-phoenix.com with your Phoenix collector endpoint

• your-api-key with your Phoenix API key

Connect via Claude Code CLI

claude mcp add phoenix npx -y @arizeai/phoenix-mcp@latest \
--baseUrl https://my-phoenix.com \
--apiKey your-api-key

Replacing:

• https://my-phoenix.com with your Phoenix collector endpoint (https://app.phoenix.arize.com default for Phoenix Cloud)
• your-api-key with your Phoenix API key

After installation, restart your Claude Code CLI session to start using the Phoenix MCP tools.

Using the Phoenix MCP Server

The MCP server can be used to interact with Prompts, Experiments, and Datasets. It can be used to retrieve information about each item, and can create and update Prompts.

Some good questions to try:

1. What prompts do I have in Phoenix?

2. Create a new prompt in Phoenix that classifies user intent

3. Update my classification prompt in Phoenix with these new options

4. Summarize the Phoenix experiments run on my agent inputs dataset

5. Visualize the results of my jailbreak dataset experiments in Phoenix

Hoping to see additional functionality?

@arizeai/phoenix-mcp is open-source! Issues and PRs welcome.

Not found