rill

The Rill MCP Server provides a structured and governed way for Large Language Models (LLMs) to interact with Rill's essential APIs, specifically focusing on consuming Rill metrics views. This ensures that LLMs can query data with predefined measures and dimensions, guaranteeing accurate and consistent responses aligned with Rill dashboards. It's primarily designed for data analysts to self-serve answers to business questions.

Key Features:

• Structured & Governed Analytics: Provides a reliable interface to your data warehouse via Rill's metrics views.
• Accurate & Consistent Data: Responses are guaranteed to be as accurate as Rill dashboards.
• AI Instructions: Allows adding ai_instructions to Rill project files (rill.yaml, metrics.yaml) to provide LLMs with additional context and guidance for best results, including how to generate Rill Explore URLs.
• Supported Actions: Exposes APIs for list_metrics_views, get_metrics_view, query_metrics_view_time_range, and query_metrics_view.

How to Use:

1. Prerequisites: Ensure you have the Rill CLI installed and a Rill Personal Access Token (if using Rill Cloud).
2. Configure Claude Desktop: Edit your claude_desktop_config.json file to include the Rill MCP server configuration under the mcpServers key.
   • For private Rill Cloud projects: Connect to https://api.rilldata.com/v1/organizations/{org}/projects/{project}/runtime/mcp with an Authorization header.
   • For public Rill Cloud projects: Connect to https://api.rilldata.com/v1/organizations/demo/projects/rill-github-analytics/runtime/mcp.
   • For locally running Rill Developer: Connect to http://localhost:9009/mcp. These configurations typically use npx mcp-remote as the client command to connect to the Rill MCP server endpoint.
3. Restart Claude Desktop: Apply changes to the configuration.
4. Enhance LLM Context: Add ai_instructions to your rill.yaml or specific metrics.yaml files to guide the LLM on how to use Rill tools and format responses, including generating Rill Explore URLs.

The Rill MCP Server empowers users to get quick, overall summaries and insights from their data through natural language conversations, leveraging the trustworthiness of Rill's existing metrics.