The dbt Cloud MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your dbt Cloud account. It provides structured and secure access to your dbt Cloud projects, jobs, runs, environments, metadata, artifacts, and Semantic Layer data, so your agent can monitor analytics work, start or manage runs, investigate failures, and query governed metrics on your behalf.
- Project and job operations: Have your agent list dbt Cloud projects and environments, inspect job settings, start job runs, cancel active runs, or retry failed runs.
- Run monitoring and troubleshooting: Let the agent find runs by job, project, environment, status, or time, inspect run details, and retrieve failure information and step logs.
- Artifact and metadata discovery: Direct your agent to list available run artifacts, download selected artifact files, explore metadata, retrieve specific resources, and inspect model health.
- Semantic Layer exploration: Instruct your agent to find governed metrics and saved queries, list available dimensions and entities, and query semantic metrics for analysis.
- Administrative resource management: Have your agent read, create, update, or delete supported dbt Cloud administrative resources through documented operations.