Brummer MCP Server

# Brummer Regression Test Suite This directory contains a comprehensive regression testing framework for Brummer that validates all core functionality in both TUI and no-TUI modes. ## Overview The test suite ensures that: - ✅ **MCP Server** works correctly in both modes - ✅ **Proxy Server** handles URL detection and request forwarding - ✅ **Logging System** captures and displays output properly - ✅ **Process Management** starts, monitors, and stops subprocesses - ✅ **Integration** between all components functions seamlessly ## Test Categories ### 🔌 MCP Tests (`mcp_tests.go`) - MCP server startup in both TUI and no-TUI modes - MCP server URL display in system messages - JSON-RPC request handling - Session tracking and connection types - Debug mode functionality ### 🌐 Proxy Tests (`proxy_tests.go`) - Proxy server startup and URL detection - HTTP request handling and forwarding - Proxy disable flag (`--no-proxy`) - Integration with process output parsing ### 📝 Logging Tests (`logging_tests.go`) - System message logging in both modes - Process output capture and display - Error detection and highlighting - Log timestamps and formatting - Log filtering capabilities (TUI) ### ⚙️ Process Tests (`process_tests.go`) - Process startup and management - Multiple concurrent processes - Process exit handling and cleanup - Process ID generation - Script detection from package.json ### 🔗 Integration Tests (`integration_tests.go`) - Full stack functionality (all components together) - MCP + Proxy integration - Process + Logging integration - Debug mode comprehensive testing - Cleanup and shutdown procedures ## Running Tests ### Quick Start ```bash # Run all regression tests make test-regression # Run with verbose output make test-regression-verbose # Run specific component tests make test-mcp make test-proxy make test-logging make test-processes make test-integration ``` ### Manual Execution ```bash # Build and run all tests ./test/run_tests.sh # Run with verbose output ./test/run_tests.sh --verbose # Use existing binary (skip build) ./test/run_tests.sh --skip-build # Run specific tests ./test/run_tests.sh --filter MCP ``` ### Advanced Usage ```bash # Use custom binary ./test/run_tests.sh --binary /path/to/custom/brum # Run only integration tests ./test/run_tests.sh --filter Integration # Verbose output with custom binary ./test/run_tests.sh --verbose --binary ./my-brum ``` ## Test Structure ### Test Workspace Each test run creates a temporary workspace (`test_workspace/`) containing: - `package.json` with test scripts - Sample configuration files - Temporary process outputs ### Test Results Tests report: - ✅ **Pass/Fail status** for each test - ⏱️ **Execution duration** - 📝 **Detailed output** and error messages - 📊 **Component-wise statistics** - 🎯 **Mode coverage** (TUI vs no-TUI) ### Sample Output ``` 🧪 Starting Brummer Regression Test Suite ========================================== 📋 Running MCP Server Tests... ✅ PASS [NoTUI] MCP - MCP Server Startup (1.2s) ✅ PASS [TUI] MCP - MCP Server Startup (3.1s) ✅ PASS [NoTUI] MCP - MCP URL Display (0.8s) ✅ PASS [NoTUI] MCP - JSON-RPC Requests (2.3s) 📋 Running Proxy Server Tests... ✅ PASS [NoTUI] Proxy - Proxy Server Startup (1.5s) ✅ PASS [TUI] Proxy - Proxy Server Startup (3.2s) ✅ PASS [NoTUI] Proxy - URL Detection (1.8s) ================================================== 📊 REGRESSION TEST SUMMARY ================================================== ✅ MCP: 8/8 passed (0 failed) ✅ Proxy: 5/5 passed (0 failed) ✅ Logging: 6/6 passed (0 failed) ✅ Processes: 6/6 passed (0 failed) ✅ Integration: 6/6 passed (0 failed) -------------------------------------------------- OVERALL: 31/31 tests passed - ✅ ALL TESTS PASSED Mode Coverage: 18 TUI tests, 13 NoTUI tests ``` ## Test Development ### Adding New Tests 1. Add test function to appropriate `*_tests.go` file 2. Register test in the component's test runner 3. Follow the `TestResult` structure for consistency ### Test Function Template ```go func (ts *TestSuite) testNewFeature() TestResult { start := time.Now() result := TestResult{ Name: "New Feature", Mode: "NoTUI", // or "TUI" Component: "ComponentName", Passed: false, } // Test implementation here cmd := exec.Command("timeout", "5s", ts.BinaryPath, "args...") output, err := cmd.CombinedOutput() result.Duration = time.Since(start) // Validation logic if /* success condition */ { result.Passed = true result.Details = []string{"Success details"} } else { result.Error = "Error description" result.Details = []string{"Debug information"} } return result } ``` ### Best Practices - ⏱️ Use timeouts to prevent hanging tests - 📝 Provide detailed error messages and debug info - 🎯 Test both TUI and no-TUI modes where applicable - 🧹 Clean up resources after tests - 📊 Include meaningful assertions and validations ## Troubleshooting ### Common Issues - **Port conflicts**: Tests use dynamic port allocation - **Timing issues**: Increase timeouts for slower systems - **Binary not found**: Ensure `make build` runs successfully - **Permission errors**: Check file permissions on test scripts ### Debug Mode Run tests with `--verbose` flag to see: - Full command outputs - Detailed test execution logs - Component-specific debug information - Timing and performance metrics ### Manual Testing For debugging specific issues: ```bash # Test MCP server manually ./brum -d test_workspace --no-tui --debug # Test TUI mode manually ./brum -d test_workspace --debug # Test specific script ./brum -d test_workspace --no-tui test ``` ## CI/CD Integration The test suite is designed for automated testing: - Exit codes: 0 = success, 1 = failure - Machine-readable output available - Configurable timeouts and retries - Minimal external dependencies ### GitHub Actions Example ```yaml - name: Run Regression Tests run: | make build make test-regression ``` This comprehensive test suite ensures Brummer's reliability across all supported modes and prevents regressions during development.