Snowplow now meets you where you are in your agentic workflow. Snowplow Assistant is an AI assistant built into Snowplow Console. The Snowplow MCP server brings the same capabilities into Claude, Cursor, Codex, and other MCP-compatible AI tools.

Together, they accelerate time to use case by letting teams design tracking, manage pipelines, and resolve data quality issues in a conversation rather than across multiple tools and team handoffs.

The Snowplow MCP server is available today. Snowplow Assistant is live in trial environments today, and rolls out to all Event Studio customers next week.

Snowplow Assistant

The Snowplow Assistant is an AI assistant embedded in Snowplow Console. Teams describe what they want to do, and the Assistant calls the relevant Snowplow APIs on their behalf.

‍

‍

The Assistant supports:

• Tracking design. Draft data structures, event specifications, and tracking plans from natural language, and search Iglu Central for reusable public schemas.
• Implementation guidance. Generate tracker code for web, mobile, and server-side platforms, with built-in awareness of the Snowplow SDKs.
• Pipeline management. Inspect pipeline status, throughput, enrichment latency, and collector configuration directly from chat.
• Troubleshooting. Investigate failed events with breakdowns by app ID, tracker version, and error type, with targeted recommendations for fixes.
• Data quality. Create and manage email and Slack alerts for schema validation failures and enrichment errors.
• Agentic context configuration. Configure attribute groups, services, and interventions through Snowplow Signals, and query real-time attributes for specific users.

The Assistant runs with each user's existing Console permissions. Any action that modifies data requires explicit confirmation before it runs. The Assistant has no access to event-level PII; it works with Snowplow configuration and metadata only.

The Assistant is opt-in. An organization admin enables it from Settings in Console, and accepts terms covering LLM usage on first activation.

Snowplow MCP server

The Snowplow MCP server is a remote Model Context Protocol server that connects AI assistants to a Snowplow Console account.

It exposes the same capabilities as the Snowplow Assistant in a customer's choice of harness and model, supporting Claude.ai, Claude Desktop, Claude Code, Codex, and Cursor. Authentication runs through Snowplow Console via OAuth, and the AI operates with the connected user's existing permissions.

Configuration instructions for each supported client are in the Snowplow MCP server documentation.

CLI MCP server

For teams managing tracking plans as code, the Snowplow CLI MCP server runs locally and connects AI tools to tracking plan files on disk. It is included in the Snowplow CLI and supports Claude Desktop, VS Code, Cursor, and other MCP-compatible clients.

‍

‍

For an overview of how Snowplow supports AI and LLM workflows, see Working with AI in the Snowplow documentation.

This release is the first in a wider program of work to make AI a first-class way of working with Snowplow. Our Product and Professional Services teams will be publishing skills for Claude, Cursor, and other AI tools over the coming weeks.

Existing customers should speak to their Customer Success Manager or Snowplow Support for help getting set up. More information for new customers will follow shortly.