Snowplow MCP server

Snowplow MCP is a remote Model Context Protocol (MCP) server that lets AI assistants interact with your Snowplow Console account using natural language.

The MCP server exposes the same set of tools as the Snowplow Assistant, but in your choice of harness and model. The server authenticates through your existing Snowplow Console login via OAuth.

Configure the MCP server

Claude.ai

Claude.ai supports remote MCP servers as custom connectors. To add Snowplow MCP:

1. Open Claude.ai and navigate to Settings.
2. Under Connectors, click Add connector.
3. Fill in the connector details:
   • Name: Snowplow MCP
   • URL: https://console.snowplowanalytics.com/api/agent/mcp
4. Expand Advanced settings and enter the OAuth Client ID: NxCcdyu13Cr4umnIYw70evvUyRXRvyWf
5. Click Add.

Claude.ai will redirect you to the Snowplow login page to authorize access.

Claude Desktop

The custom connector method also works for Claude Desktop. But in case you don't have access to custom connectors, you can configure it directly in Claude Desktop using mcp-remote

Add the following to your Claude Desktop configuration filej:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

json

{
  "mcpServers": {
    "snowplow-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://console.snowplowanalytics.com/api/agent/mcp",
        "3334",
        "--static-oauth-client-info",
        "{\"client_id\":\"NxCcdyu13Cr4umnIYw70evvUyRXRvyWf\"}"
      ]
    }
  }
}

Restart Claude Desktop after saving. On startup Claude Desktop opens a browser window for Snowplow login and caches the token locally for subsequent sessions.

Claude Code

Add a remote MCP server to Claude Code by running:

bash

claude mcp add snowplow-mcp \
  --transport http \
  --client-id NxCcdyu13Cr4umnIYw70evvUyRXRvyWf \
  --callback-port 3334 \
  https://console.snowplowanalytics.com/api/agent/mcp

Run the "mcp" command, select the "snowplow-mcp" and then "Authenticate". This should open a browser page for Snowplow login.

Codex CLI

Add a remote MCP server to Codex CLI by running:

bash

codex mcp add snowplow-mcp -- \
  npx mcp-remote \
  https://console.snowplowanalytics.com/api/agent/mcp \
  3334 \
  --static-oauth-client-info \
  '{"client_id":"NxCcdyu13Cr4umnIYw70evvUyRXRvyWf"}'

Codex will redirect you to the Snowplow login page to authorize access.

Codex UI

1. Open Codex and navigate to Settings > MCP servers > Add server > STDIO.
2. Fill in the server details:
   • Name: Snowplow MCP
   • Command to launch: npx
   • Arguments (one per row):
     1. mcp-remote
     2. https://console.snowplowanalytics.com/api/agent/mcp
     3. 3334
     4. --static-oauth-client-info
     5. {"client_id":"NxCcdyu13Cr4umnIYw70evvUyRXRvyWf"}
3. Click Save.

Codex will redirect you to the Snowplow login page to authorize access.

Cursor

1. Open Cursor and navigate to Settings > Cursor Settings > Tools & MCPs.
2. Click New MCP Server. This opens the mcp.json configuration file.
3. Add the following configuration:

json

{
  "mcpServers": {
    "snowplow-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://console.snowplowanalytics.com/api/agent/mcp",
        "3334",
        "--static-oauth-client-info",
        "{\"client_id\":\"NxCcdyu13Cr4umnIYw70evvUyRXRvyWf\"}"
      ]
    }
  }
}

4. Save the file and return to Tools & MCPs. The server should appear in the list.

Cursor will redirect you to the Snowplow login page to authorize access.

Authentication

The MCP server authenticates via OAuth using your Snowplow Console credentials. The AI assistant operates with the same permissions as your user account — it can access and modify only things that you can.

The server automatically connects to the organization your account belongs to. Selecting a different organization is not currently supported.

Capabilities of the server

The MCP server gives your AI assistant read and write access to the main areas of Snowplow Console. The full list of tools is visible in your MCP client after connecting.

• Data structures and schemas — browse, inspect, and create data structures, and search Iglu Central for reusable public schemas.
• Event specifications and tracking plans — manage event specifications and tracking plans, and view event volume metrics.
• Source applications — manage source applications and their associated entities.
• Pipelines — inspect pipeline configuration, health metrics, and collector settings, including Snowplow Micro instances.
• Failed events and data quality — investigate failed events and manage data quality alerts.
• Enrichments — view and update enrichment configurations.
• Data catalog — browse and search tracked data structures.
• Signals — manage the full Signals workflow, including attribute groups, services, interventions, and publishing to compute engines.
• Documentation — fetch pages from the Snowplow documentation site for quick reference.

Example prompts

After connecting, try asking your assistant:

• "What pipelines do I have and what's their current status?"
• "Show me the failed events from the last 24 hours"
• "Create a new event specification for a signup_completed event"
• "What enrichments are enabled on my production pipeline?"
• "Search the data catalog for anything related to ecommerce"