dbt-doctor

Overview

dbt-doctor is a specialized MCP server designed to enhance dbt project management through AI-assisted quality control and governance. It provides deep context about dbt project health, allowing users to interact with AI assistants for tasks like auditing coverage, profiling data models, detecting schema drift, and auto-generating documentation. The server operates as a complement to the official dbt-labs/dbt-mcp, focusing on analytics rather than command execution. The server's architecture consists of a read-only analysis layer that connects the AI assistant with dbt projects and data warehouses, combined with a secure generation toolkit that seamlessly merges changes into existing YAML schemas. It emphasizes security through read-only SQL execution, strict SQL validation, stateless connections, and preview-before-commit mechanisms for document generation, ensuring users maintain full control over their dbt configurations.

Try asking AI

After installing, here are 3 things you can ask your AI assistant:

• list_models

  Overview of all models and their coverage status

  Try:What models are in my dbt project?Try:Show me all models with low coverage
• get_model_details

  Detailed model information including SQL, columns, lineage, and tests

  Try:Show me details for the fct_orders modelTry:What tests are defined for the stg_customers model?
• audit_project

  Project health score and naming convention violations

  Try:What's the health score of my dbt project?Try:Find any naming convention violations
• check_test_coverage

  Models ranked by their test coverage percentage

  Try:Which models have the lowest test coverage?Try:List all models with 0% coverage
• analyze_dag

  Detection of orphan models and high fan-out nodes

  Try:Are there any orphan models in my project?Try:Find models with high fan-out
• get_project_health

  Single-call dashboard summarizing project status

  Try:Give me a complete health report for my projectTry:Show me project health dashboard
• profile_model

  Batched column statistics for a model

  Try:Profile the fct_orders modelTry:Get column statistics for stg_customers
• execute_query

  Read-only SQL execution against your warehouse

  Try:Count total orders in the last monthTry:Show top 10 customers by revenue
• detect_schema_drift

  Comparison of database columns against manifest definitions

  Try:Check for schema drift in my modelsTry:Find any columns not in manifest.json
• suggest_tests

  Translation of profile data into dbt test recommendations

  Try:Suggest tests for the fct_orders modelTry:What tests should I add based on profiling?
• update_model_yaml

  Safe merging of documentation and tests to schema.yml

  Try:Update schema.yml with new testsTry:Add documentation to the stg_customers model
• generate_model_docs

  Complete end-to-end documentation workflow

  Try:Generate documentation for all modelsTry:Auto-document models with low coverage

Comparable tools

Installation

pip install dbt-doctor

{
  "mcpServers": {
    "dbt-doctor": {
      "command": "dbt-doctor",
      "args": ["--project-dir", "/absolute/path/to/your/dbt/project"]
    }
  }
}

{
  "mcpServers": {
    "dbt-doctor": {
      "command": "dbt-doctor",
      "args": ["--project-dir", "/absolute/path/to/your/dbt/project"]
    }
  }
}

Compare dbt-doctor with