Databricks MCP Server

Instructions

Implementation Reference

• src/server/databricks_mcp_server.py:145-156 (handler)

  MCP tool handler for 'list_notebooks': decorator registers the tool and the function wraps the API call to list notebooks in a Databricks workspace directory, handling params and errors.

      name="list_notebooks",
      description="List notebooks in a workspace directory with parameter: path (required)",
  )
  async def list_notebooks(params: Dict[str, Any]) -> List[TextContent]:
      logger.info(f"Listing notebooks with params: {params}")
      try:
          result = await notebooks.list_notebooks(params.get("path"))
          return [{"text": json.dumps(result)}]
      except Exception as e:
          logger.error(f"Error listing notebooks: {str(e)}")
          return [{"text": json.dumps({"error": str(e)})}]

• src/api/notebooks.py:93-107 (helper)

  Helper function implementing the core logic: calls Databricks /api/2.0/workspace/list API to retrieve the list of notebooks/files in the specified path.

  async def list_notebooks(path: str) -> Dict[str, Any]:
      """
      List notebooks in a workspace directory.
      
      Args:
          path: The path to list
          
      Returns:
          Response containing the directory listing
          
      Raises:
          DatabricksAPIError: If the API request fails
      """
      logger.info(f"Listing notebooks in path: {path}")
      return make_api_request("GET", "/api/2.0/workspace/list", params={"path": path})

• src/server/databricks_mcp_server.py:145-148 (registration)

  Tool registration via @self.tool decorator specifying name and description (input schema inferred).

      name="list_notebooks",
      description="List notebooks in a workspace directory with parameter: path (required)",
  )
  async def list_notebooks(params: Dict[str, Any]) -> List[TextContent]: