Databricks MCP Server

# Databricks MCP Tools - Modular Structure This directory contains the modularized MCP tools for Databricks operations. The original `tools.py` file (4231 lines) has been broken down into logical, manageable modules. ## Structure ``` server/tools/ ├── __init__.py # Main entry point that imports and registers all tools ├── core.py # Core/health tools (1 tool) ├── sql_operations.py # SQL warehouse and query management (15 tools) ├── unity_catalog.py # Unity Catalog operations (20 tools) ├── data_management.py # DBFS, volumes, and data operations (10 tools) ├── jobs_pipelines.py # Job and pipeline management (20 tools) ├── workspace_files.py # Workspace file operations (5 tools) ├── dashboards.py # Dashboard and monitoring tools (8 tools) ├── repositories.py # Git repository management (10 tools) ├── governance.py # Governance rules and data lineage (15 tools) └── test_imports.py # Test file for verifying imports ``` ## Tool Distribution | Module | Tools | Description | |--------|-------|-------------| | **core.py** | 1 | Basic health checks and core functionality | | **sql_operations.py** | 15 | SQL warehouse management, query execution, and monitoring | | **unity_catalog.py** | 20 | Catalog, schema, table, and metadata operations | | **data_management.py** | 10 | DBFS operations, external locations, storage credentials | | **jobs_pipelines.py** | 20 | Job and DLT pipeline management | | **workspace_files.py** | 5 | Workspace file and directory operations | | **dashboards.py** | 8 | Lakeview and legacy dashboard management | | **repositories.py** | 10 | Git repository operations and branch management | | **governance.py** | 15 | Audit logs, governance rules, and data lineage | **Total: 104 tools** organized into 9 logical modules ## Benefits of Modularization 1. **Maintainability**: Each module focuses on a specific domain, making code easier to understand and maintain 2. **Readability**: Smaller files are easier to navigate and debug 3. **Collaboration**: Multiple developers can work on different modules simultaneously 4. **Testing**: Individual modules can be tested in isolation 5. **Scalability**: New tools can be added to appropriate modules without cluttering the main file 6. **Documentation**: Each module has clear purpose and can be documented independently ## Usage The main `load_tools()` function in `__init__.py` automatically imports and registers all tools from each module. No changes are needed in the calling code - the interface remains the same. ## Adding New Tools To add new tools: 1. Identify the appropriate module based on functionality 2. Add the tool function to that module 3. Ensure the tool is properly decorated with `@mcp_server.tool` 4. The tool will automatically be available when the module is loaded ## Migration Notes - Original `tools.py` has been backed up as `tools.py.backup` - All existing functionality is preserved - No breaking changes to the API - Tools are organized by domain rather than alphabetically for better logical grouping