pawrly

Documentation

Pawrly documentation

Everything to install Pawrly, connect your sources, model your data, and query it with SQL — from the CLI, MCP, or an agent.

Start here

OverviewWhat Pawrly is and how the pieces fit together.CLIEvery command: sql, schema, validate, serve, status, mcp-stdio.ConfigurationThe pawrly.yaml schema — sources, secrets, caching, safety.

Connect data

SourcesWire up APIs, files, databases, warehouses, lakehouses, and MCP servers.

Model & serve

Semantic layerDimensions, measures, relationships, segments, RLS, and pre-aggregations.Materialized tablesPin a query, file, or URL as a self-backed table.

For agents

MCP serverRun Pawrly as an MCP server and consume other MCP servers as sources.

Operate

ObservabilityTraces, metrics, and a queryable activity log.ArchitectureEngine internals and local vs daemon behavior.ConsoleThe web console for browsing sources and running queries.