Skip to content

Mostly Right platform docs

Use datasets maintained by Mostly Right, or create one with agents that find the sources, prepare the data, and configure how it stays updated.

Browse the public catalog before building something new. Each dataset has a documented schema, freshness information, and a stable page where you can inspect the available tables.

Download the full Parquet file when you need the complete history. Use the Public Dataset API for bounded JSON queries, or connect an AI tool through MCP.

Describe the data you need. Agents find and compare sources, configure the ingest steps, clean and join the data, and suggest an update schedule based on when the sources actually publish.

The Python and TypeScript packages are a separate legacy product with their own installation guides, API reference, and release history.

Open the SDK docs