Bruin

Bruin is a data platform for building, scheduling, and monitoring end-to-end data pipelines across warehouses like BigQuery, Snowflake, Databricks, Redshift, Postgres, and DuckDB. This connector gives Claude direct access to your Bruin Cloud workspace so you can operate your pipelines in plain language instead of clicking through the UI. Ask Claude to list your pipelines and assets, page through run history, and pull the execution logs for a failed step so you can diagnose a break without leaving the conversation. Inspect an asset’s health timeline interval by interval, surface pipeline validation errors, and break down warehouse spend by pipeline, asset, user, or project to find what’s driving cost. When you know what needs to happen, Claude can act: trigger a pipeline run for a given data interval, rerun a failed run or a single asset with its upstream or downstream dependencies, launch and manage backfills across a date range, mark runs or assets as success/failed, annotate runs with notes, pause or enable a pipeline, and create or remove warehouse connections. Every state-changing tool is annotated as destructive, so Claude asks before it acts. Connection credentials are never returned - the connector exposes connection names, types, and timestamps only, and never secret values. Access is scoped to the Bruin team you pick during the OAuth consent step and to your own role’s permissions in that team.