Igloo MCP

# igloo-mcp Features Overview ## 🚀 Core Features ### 1. SQL Execution & Query Management **Description**: Direct SQL execution against Snowflake with structured output and error handling. **Capabilities**: - Execute arbitrary SQL queries with configurable output formats (JSON, CSV) - Context-aware execution with warehouse/database/schema/role overrides - Query result parsing and structured output - Timeout and error handling with detailed diagnostics **MCP Tools**: `execute_query`, `preview_table` **AI Assistant Usage**: ``` "Execute this query: SELECT * FROM CUSTOMERS LIMIT 10" "Run the SQL query in query.sql file" "Preview the CUSTOMERS table" ``` **Testing Coverage**: ✅ **WELL COVERED** - Unit tests for SnowCLI wrapper (`test_snow_cli.py`) - Query parsing and output formatting tests - Error handling scenarios - Mock-based testing for different output formats --- ### 2. Data Catalog Generation **Description**: Comprehensive metadata extraction and cataloging of Snowflake databases. **Capabilities**: - Parallel metadata extraction from INFORMATION_SCHEMA and SHOW commands - Support for tables, views, materialized views, functions, procedures, tasks, dynamic tables - DDL extraction with concurrency control - JSON/JSONL output formats with configurable structure - Incremental catalog updates with change tracking - Cross-database catalog generation **MCP Tools**: `build_catalog`, `get_catalog_summary` **AI Assistant Usage**: ``` "Build a catalog for the ANALYTICS database" "Create a comprehensive catalog with DDL for all databases" "Show me the catalog summary" ``` **Testing Coverage**: ⚠️ **PARTIALLY COVERED** - ✅ Basic catalog building functionality - ✅ Service layer architecture (`test_services.py`) - ❌ **MISSING**: Large-scale catalog testing - ❌ **MISSING**: DDL extraction testing - ❌ **MISSING**: Incremental update testing - ❌ **MISSING**: Cross-database catalog testing --- ### 3. Data Lineage Analysis **Description**: Build and analyze data lineage graphs to understand data flow and dependencies. **Capabilities**: - SQL parsing with SQLGlot for dependency extraction - Bidirectional lineage traversal (upstream/downstream) - Multiple output formats (text, JSON, HTML, DOT) - Cross-database lineage resolution - Lineage caching for performance - Interactive HTML visualizations **MCP Tools**: `query_lineage` **AI Assistant Usage**: ``` "Show me lineage for ANALYTICS.CUSTOMERS going both upstream and downstream" "Trace data lineage with depth 3" "Generate an HTML visualization of lineage for MY_TABLE" ``` **Testing Coverage**: ✅ **WELL COVERED** - Comprehensive lineage tests (`test_lineage.py`, `test_advanced_lineage.py`) - SQL parsing and dependency extraction - Graph traversal algorithms - Multiple output format testing - Edge case handling --- ### 4. Dependency Graph Generation **Description**: Create visual dependency graphs showing object relationships. **Capabilities**: - Object relationship mapping from metadata - DOT format output for Graphviz rendering - JSON output for programmatic consumption - Circular dependency detection - Hierarchical visualization support **MCP Tools**: `build_dependency_graph` **AI Assistant Usage**: ``` "Build a dependency graph for the ANALYTICS database" "Create a dependency graph in DOT format" "Show me object dependencies as a graph" ``` **Testing Coverage**: ⚠️ **PARTIALLY COVERED** - ✅ Basic dependency graph construction - ❌ **MISSING**: Complex relationship testing - ❌ **MISSING**: Circular dependency detection testing - ❌ **MISSING**: Large graph performance testing --- ### 5. Configuration Management & Profile Validation (v2.0.0) **Description**: Robust configuration system with advanced profile validation and health monitoring. **Core Capabilities**: - YAML-based configuration files - Environment variable overrides - Multiple Snowflake profile support - Default value cascading - Configuration validation **Profile Validation Features**: - **Startup validation**: Profile issues detected before server becomes available - **Clear error messages**: No more confusing timeout errors - **MCP-compliant error responses**: Structured error format with specific error codes - **Real-time diagnostics**: Health monitoring tools for ongoing validation - **Actionable guidance**: Specific next steps for fixing configuration issues - **Profile health caching**: Efficient validation with TTL-based caching **MCP Server Usage**: ```bash # Start MCP server with profile validation (v2.0+) igloo-mcp # Shows validation success/failure immediately export SNOWFLAKE_PROFILE=my-profile # Clear profile selection # Profile selection via command line igloo-mcp --profile my-profile ``` **MCP Tools**: - `health_check`: Comprehensive server health status - `check_profile_config`: Profile validation and diagnostics - `get_resource_status`: Resource availability checking - `check_resource_dependencies`: Dependency validation **Error Handling Improvements**: **Benefits**: - Immediate, specific error messages with context - No generic timeout errors **Example Enhanced Error Response**: ```json { "error": { "code": -32004, "message": "Snowflake profile validation failed", "data": { "profile_name": "default", "available_profiles": ["dev", "prod"], "suggestion": "Set SNOWFLAKE_PROFILE environment variable" } } } ``` **Testing Coverage**: ✅ **EXCELLENT COVERAGE** - Configuration loading and validation (`test_config.py`) - Environment variable handling - YAML serialization/deserialization - Profile management - Health monitoring tests (`test_mcp_health.py`) - Profile validation tests - MCP error response testing --- ### 6. MCP Server Integration (v2.0.0) **Description**: Model Context Protocol server for AI assistant integration with advanced health monitoring and reliability. **Core Capabilities**: - JSON-RPC 2.0 protocol implementation - Tool-based interface for AI assistants - Async operation support - VS Code, Cursor, Claude Code compatibility - Secure authentication through existing Snowflake CLI profiles **Reliability Features**: - **Proactive validation**: Profile validation during server startup lifecycle - **Circuit breaker pattern**: Fault-tolerant Snowflake operations - **Health monitoring**: Real-time component health tracking - **Structured error responses**: MCP-compliant error codes and context - **Resource management**: Dependency tracking and availability monitoring - **Graceful degradation**: Partial functionality when components fail **CLI Usage**: ```bash # Startup with validation igloo-mcp # Shows immediate validation feedback # Expected successful startup: # ✓ Snowflake profile validation successful: dev # ✓ Profile health check passed for: dev # ✓ Snowflake connection health check passed # Starting FastMCP server using transport=stdio ``` **MCP Tools**: - **Core tools**: All existing tools (execute_query, build_catalog, etc.) - **Diagnostic tools**: health_check, test_connection **Reliability Infrastructure**: - **MCPHealthMonitor**: Comprehensive health status tracking - **MCPResourceManager**: Resource dependency management - **Error categorization**: Connection, Permission, Timeout, Configuration errors - **Performance optimization**: Caching with TTL for health checks **Testing Coverage**: ✅ **EXCELLENT COVERAGE** - MCP server functionality (`test_mcp_server.py`) - Tool registration and execution - Error handling and response formatting - Mock-based testing for external dependencies - Health monitoring system tests - Circuit breaker pattern tests - Resource management tests - Profile validation integration tests --- ## 🔧 Advanced Features ### 7. Column-Level Lineage (Advanced) **Description**: Track data flow at individual column granularity. **Capabilities**: - SQL parsing for column-level dependencies - Transformation type detection (DIRECT, FUNCTION, AGGREGATE, etc.) - Confidence scoring for transformations - Source-to-target column mapping **Testing Coverage**: ✅ **WELL COVERED** - Advanced lineage tests cover column parsing - SQL transformation detection - Complex query analysis --- ### 8. Cross-Database Lineage (Advanced) **Description**: Unified lineage analysis across multiple databases. **Capabilities**: - Multi-database catalog merging - Cross-database reference resolution - Database hub detection - Boundary analysis **Testing Coverage**: ⚠️ **PARTIALLY COVERED** - ✅ Basic cross-database functionality - ❌ **MISSING**: Multi-database integration testing - ❌ **MISSING**: Large-scale cross-database scenarios --- ### 9. Impact Analysis (Advanced) **Description**: Analyze potential impact of changes before implementation. **Capabilities**: - Blast radius calculation - Change impact severity scoring - Single point of failure detection - Propagation time analysis **Testing Coverage**: ✅ **WELL COVERED** - Impact analysis algorithms tested - Circular dependency detection - Missing node handling --- ### 10. Time-Travel Lineage (Advanced) **Description**: Track lineage evolution over time with snapshots. **Capabilities**: - Lineage snapshot capture - Historical comparison - Evolution tracking - Pattern detection **Testing Coverage**: ✅ **WELL COVERED** - Snapshot comparison functionality - Time-travel scenarios tested --- ### 11. External Source Integration (Advanced) **Description**: Map external data sources including cloud storage. **Capabilities**: - S3, Azure Blob, GCS integration - Stage configuration tracking - Security analysis - Access pattern monitoring **Testing Coverage**: ⚠️ **PARTIALLY COVERED** - ✅ Basic external source mapping - ❌ **MISSING**: Full credential handling tests (1 failing test) - ❌ **MISSING**: Integration with actual cloud storage --- ## 🛡️ Infrastructure Features ### 12. Circuit Breaker Pattern **Description**: Prevent cascade failures with intelligent failure handling. **Capabilities**: - Configurable failure thresholds - Exponential backoff - State management (closed, open, half-open) - Decorator support for easy application **Testing Coverage**: ✅ **EXCELLENT COVERAGE** - Comprehensive circuit breaker tests (`test_circuit_breaker.py`) - All states and transitions tested - Recovery scenarios covered - Error categorization testing --- ### 13. Health Monitoring **Description**: System health monitoring and diagnostics. **Capabilities**: - Connection health checks - Circuit breaker status reporting - System metrics collection - Health endpoint for monitoring **Testing Coverage**: ✅ **WELL COVERED** - Health status reporting (`test_services.py`) - Circuit breaker integration - Error state handling --- ### 14. Comprehensive Error Handling **Description**: Structured error handling with categorization and context. **Capabilities**: - Error categorization (Connection, Permission, Timeout) - Context-aware error reporting - Error aggregation for batch operations - Safe execution patterns with fallbacks **Testing Coverage**: ✅ **EXCELLENT COVERAGE** - Error handling strategies (`test_error_handling.py`) - Error categorization testing - Context preservation - Fallback mechanism testing --- ## 📊 Testing Coverage Summary ### Overall Test Statistics - **Total Tests**: 80+ passing tests - **Test Files**: 7 primary test files - **Coverage Estimate**: ~25-30% of codebase ### Coverage by Category | Feature Category | Coverage Level | Test Quality | Missing Areas | |------------------|----------------|--------------|---------------| | **Core CLI** | ✅ Excellent | High | Minor edge cases | | **Configuration** | ✅ Excellent | High | None significant | | **MCP Server** | ✅ Excellent | High | Integration scenarios | | **Basic Lineage** | ✅ Excellent | High | Large graph performance | | **Infrastructure** | ✅ Excellent | High | None significant | | **Data Catalog** | ⚠️ Partial | Medium | Large-scale, DDL, incremental | | **Dependency Graphs** | ⚠️ Partial | Medium | Complex relationships | | **Advanced Lineage** | ✅ Good | Medium-High | Multi-database scenarios | | **External Sources** | ⚠️ Partial | Medium | Cloud integration | ### High-Priority Testing Gaps #### 1. **Catalog Testing** (Priority: HIGH) ```python # MISSING: Large catalog testing def test_large_catalog_performance(): """Test catalog building with 1000+ objects""" # MISSING: DDL extraction testing def test_ddl_extraction_concurrent(): """Test concurrent DDL fetching""" # MISSING: Incremental updates def test_incremental_catalog_updates(): """Test incremental catalog building""" ``` #### 2. **Integration Testing** (Priority: HIGH) ```python # MISSING: End-to-end MCP workflows def test_mcp_end_to_end_workflow(): """Test complete AI assistant workflow""" # MISSING: Cross-database scenarios def test_cross_database_lineage_integration(): """Test lineage across multiple databases""" ``` #### 3. **Performance Testing** (Priority: MEDIUM) ```python # MISSING: Large dataset handling def test_large_graph_performance(): """Test lineage with 10,000+ objects""" # MISSING: Memory usage testing def test_memory_usage_large_catalogs(): """Monitor memory usage during catalog building""" ``` #### 4. **External Integration Testing** (Priority: MEDIUM) ```python # MISSING: Cloud storage integration def test_s3_integration(): """Test S3 bucket mapping and access""" # MISSING: Credential security def test_credential_security(): """Test credential handling security""" ``` ### Recommended Testing Improvements #### 1. **Add Property-Based Testing** ```python from hypothesis import given, strategies as st @given(st.text(), st.integers(min_value=1, max_value=100)) def test_catalog_building_properties(database_name, object_count): """Property-based testing for catalog building""" ``` #### 2. **Add Load Testing** ```python def test_concurrent_mcp_requests(): """Test MCP server under concurrent load""" def test_large_lineage_graph_traversal(): """Test lineage traversal with large graphs""" ``` #### 3. **Add Integration Scenarios** ```python def test_full_data_pipeline_analysis(): """Test complete data pipeline analysis workflow""" def test_impact_analysis_real_world(): """Test impact analysis with realistic scenarios""" ``` #### 4. **Add Error Recovery Testing** ```python def test_circuit_breaker_recovery_scenarios(): """Test various circuit breaker recovery patterns""" def test_partial_failure_handling(): """Test handling of partial failures in batch operations""" ``` ## 🎯 Testing Strategy Recommendations ### Short Term (Next Sprint) 1. **Fix failing external source test** - Address credential handling issue 2. **Add catalog performance tests** - Test with 1000+ objects 3. **Add MCP integration tests** - End-to-end workflows 4. **Add error recovery tests** - Circuit breaker scenarios ### Medium Term (Next Month) 1. **Property-based testing** - Add hypothesis-based tests 2. **Load testing** - Concurrent operations and large datasets 3. **Security testing** - Credential handling and injection prevention 4. **Memory profiling** - Monitor resource usage ### Long Term (Next Quarter) 1. **Automated performance benchmarks** - Regression detection 2. **Chaos engineering** - Fault injection testing 3. **Integration test suite** - Real Snowflake environments 4. **Documentation testing** - Verify all examples work ## 🎉 Conclusion igloo-mcp provides a comprehensive suite of features for Snowflake data management, with particularly strong coverage in: - Core CLI functionality - Circuit breaker and reliability patterns - Error handling and monitoring - Basic lineage analysis - MCP server integration The testing coverage is solid for core features (~80 passing tests), with room for improvement in: - Large-scale catalog operations - Complex dependency scenarios - External system integrations - Performance and load testing The codebase demonstrates excellent software engineering practices with the recent addition of circuit breakers, proper error handling, and comprehensive service layer architecture.