CHANGELOG.md•86.4 kB
# Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [Unreleased] - Phase 0: Connection Operations Critical Fixes
### Fixed
- **🐛 CRITICAL: Fixed `addConnection` sourceIndex handling (Issue #272, discovered in hands-on testing)**
- Multi-output nodes (IF, Switch) now work correctly with sourceIndex parameter
- Changed from `||` to `??` operator to properly handle explicit 0 values
- Added defensive array validation before accessing indices
- Improves rating from 3/10 to 8/10 for multi-output node scenarios
- **Impact**: IF nodes, Switch nodes, and all conditional routing now reliable
- **🐛 CRITICAL: Added runtime validation for `updateConnection` (Issue #272, #204)**
- Prevents server crashes when `updates` object is missing
- Provides helpful error message with:
- Clear explanation of what's wrong
- Correct format example
- Suggestion to use removeConnection + addConnection for rewiring
- Validates `updates` is an object, not string or other type
- **Impact**: No more cryptic "Cannot read properties of undefined" crashes
### Enhanced
- **Error Messages**: `updateConnection` errors now include actionable guidance
- Example format shown in error
- Alternative approaches suggested (removeConnection + addConnection)
- Clear explanation that updateConnection modifies properties, not targets
### Testing
- Added 8 comprehensive tests for Phase 0 fixes
- 2 tests for updateConnection validation (missing updates, invalid type)
- 5 tests for sourceIndex handling (IF nodes, parallel execution, Switch nodes, explicit 0)
- 1 test for complex multi-output routing scenarios
- All 126 existing tests still passing
### Documentation
- Updated tool documentation to clarify:
- `addConnection` now properly handles sourceIndex (Phase 0 fix noted)
- `updateConnection` REQUIRES 'updates' object (Phase 0 validation noted)
- Added pitfalls about updateConnection limitations
- Clarified that updateConnection modifies properties, NOT connection targets
### Developer Experience
- More defensive programming throughout connection operations
- Better use of nullish coalescing (??) vs. logical OR (||)
- Clear inline comments explaining expected behavior
- Improved type safety with runtime guards
### References
- Comprehensive analysis: `docs/local/connection-operations-deep-dive-and-improvement-plan.md`
- Based on hands-on testing with n8n-mcp-tester agent
- Overall experience rating improved from 4.5/10 to estimated 6/10
## [2.14.4] - 2025-09-30
### Added
- **Workflow Cleanup Operations**: Two new operations for `n8n_update_partial_workflow` to handle broken workflow recovery
- `cleanStaleConnections`: Automatically removes all connections referencing non-existent nodes
- Essential after node renames or deletions that leave dangling connection references
- Supports `dryRun: true` mode to preview what would be removed
- Removes both source and target stale connections
- `replaceConnections`: Replace entire connections object in a single operation
- Faster than crafting many individual connection operations
- Useful for bulk connection rewiring
- **Graceful Error Handling for Connection Operations**: Enhanced `removeConnection` operation
- New `ignoreErrors` flag: When `true`, operation succeeds even if connection doesn't exist
- Perfect for cleanup scenarios where you're not sure if connections exist
- Maintains backwards compatibility (defaults to `false` for strict validation)
- **Best-Effort Mode**: New `continueOnError` mode for `WorkflowDiffRequest`
- Apply valid operations even if some fail
- Returns detailed results with `applied` and `failed` operation indices
- Breaks atomic guarantees intentionally for bulk cleanup scenarios
- Maintains atomic mode as default for safety
### Enhanced
- **Tool Documentation**: Updated `n8n_update_partial_workflow` documentation
- Added examples for cleanup scenarios
- Documented new operation types and modes
- Added best practices for workflow recovery
- Clarified atomic vs. best-effort behavior
- **Type System**: Extended workflow diff types
- Added `CleanStaleConnectionsOperation` interface
- Added `ReplaceConnectionsOperation` interface
- Extended `WorkflowDiffResult` with `applied`, `failed`, and `staleConnectionsRemoved` fields
- Updated type guards for new connection operations
### Testing
- Added comprehensive test suite for v2.14.4 features
- 15 new tests covering all new operations and modes
- Tests for cleanStaleConnections with various stale scenarios
- Tests for replaceConnections validation
- Tests for ignoreErrors flag behavior
- Tests for continueOnError mode with mixed success/failure
- Backwards compatibility verification tests
### Impact
- **Time Saved**: Reduces broken workflow fix time from 10-15 minutes to 30 seconds
- **Token Efficiency**: `cleanStaleConnections` is 1 operation vs 10+ manual operations
- **User Experience**: Dramatically improved workflow recovery capabilities
- **Backwards Compatibility**: 100% - all additions are optional and default to existing behavior
## [2.13.2] - 2025-01-24
### Added
- **Operation and Resource Validation with Intelligent Suggestions**: New similarity services for n8n node configuration validation
- `OperationSimilarityService`: Validates operations and suggests similar alternatives using Levenshtein distance and pattern matching
- `ResourceSimilarityService`: Validates resources with automatic plural/singular conversion and typo detection
- Provides "Did you mean...?" suggestions when invalid operations or resources are used
- Example: `operation: "listFiles"` suggests `"search"` for Google Drive nodes
- Example: `resource: "files"` suggests singular `"file"` with 95% confidence
- Confidence-based suggestions (minimum 30% threshold) with contextual fix messages
- Resource-aware operation filtering ensures suggestions are contextually appropriate
- 5-minute cache duration for performance optimization
- Integrated into `EnhancedConfigValidator` for seamless validation flow
- **Custom Error Handling**: New `ValidationServiceError` class for better error management
- Proper error chaining with cause tracking
- Specialized factory methods for common error scenarios
- Type-safe error propagation throughout the validation pipeline
### Enhanced
- **Code Quality and Security Improvements** (based on code review feedback):
- Safe JSON parsing with try-catch error boundaries
- Type guards for safe property access (`getOperationValue`, `getResourceValue`)
- Memory leak prevention with periodic cache cleanup
- Performance optimization with early termination for exact matches
- Replaced magic numbers with named constants for better maintainability
- Comprehensive JSDoc documentation for all public methods
- Improved confidence calculation for typos and transpositions
### Fixed
- **Test Compatibility**: Updated test expectations to correctly handle exact match scenarios
- **Cache Management**: Fixed cache cleanup to prevent unbounded memory growth
- **Validation Deduplication**: Enhanced config validator now properly replaces base validator errors with detailed suggestions
### Testing
- Added comprehensive test coverage for similarity services (37 new tests)
- All unit tests passing with proper edge case handling
- Integration confirmed via n8n-mcp-tester agent validation
## [2.13.1] - 2025-01-24
### Changed
- **Removed 5-operation limit from n8n_update_partial_workflow**: The workflow diff engine now supports unlimited operations per request
- Previously limited to 5 operations for "transactional integrity"
- Analysis revealed the limit was unnecessary - the clone-validate-apply pattern already ensures atomicity
- All operations are validated before any are applied, maintaining data integrity
- Enables complex workflow refactoring in single API calls
- Updated documentation and examples to demonstrate large batch operations (26+ operations)
## [2.13.0] - 2025-01-24
### Added
- **Webhook Path Autofixer**: Automatically generates UUIDs for webhook nodes missing path configuration
- Generates unique UUID for both `path` parameter and `webhookId` field
- Conditionally updates typeVersion to 2.1 only when < 2.1 to ensure compatibility
- High confidence fix (95%) as UUID generation is deterministic
- Resolves webhook nodes showing "?" in the n8n UI
- **Enhanced Node Type Suggestions**: Intelligent node type correction with similarity matching
- Multi-factor scoring system: name similarity, category match, package match, pattern match
- Handles deprecated package prefixes (n8n-nodes-base. → nodes-base.)
- Corrects capitalization mistakes (HttpRequest → httpRequest)
- Suggests correct packages (nodes-base.openai → nodes-langchain.openAi)
- Only auto-fixes suggestions with ≥90% confidence
- 5-minute cache for performance optimization
- **n8n_autofix_workflow Tool**: New MCP tool for automatic workflow error correction
- Comprehensive documentation with examples and best practices
- Supports 5 fix types: expression-format, typeversion-correction, error-output-config, node-type-correction, webhook-missing-path
- Confidence-based system (high/medium/low) for safe fixes
- Preview mode to review changes before applying
- Integrated with workflow validation pipeline
### Fixed
- **Security**: Eliminated ReDoS vulnerability in NodeSimilarityService
- Replaced all regex patterns with string-based matching
- No performance impact while maintaining accuracy
- **Performance**: Optimized similarity matching algorithms
- Levenshtein distance algorithm optimized from O(m*n) space to O(n)
- Added early termination for performance improvement
- Cache invalidation with version tracking prevents memory leaks
- **Code Quality**: Improved maintainability and type safety
- Extracted magic numbers into named constants
- Added proper type guards for runtime safety
- Created centralized node-type-utils for consistent type normalization
- Fixed silent failures in setNestedValue operations
### Changed
- Template sanitizer now includes defensive null checks for runtime safety
- Workflow validator uses centralized type normalization utility
## [2.12.2] - 2025-01-22
### Changed
- Updated n8n dependencies to latest versions:
- n8n: 1.111.0 → 1.112.3
- n8n-core: 1.110.0 → 1.111.0
- n8n-workflow: 1.108.0 → 1.109.0
- @n8n/n8n-nodes-langchain: 1.110.0 → 1.111.1
- Rebuilt node database with 536 nodes (438 from n8n-nodes-base, 98 from langchain)
## [2.12.1] - 2025-01-21
### Added
- **Comprehensive Expression Format Validation System**: Three-tier validation strategy for n8n expressions
- **Universal Expression Validator**: 100% reliable detection of expression format issues
- Enforces required `=` prefix for all expressions `{{ }}`
- Validates expression syntax (bracket matching, empty expressions)
- Detects common mistakes (template literals, nested brackets, double prefixes)
- Provides confidence score of 1.0 for universal rules
- **Confidence-Based Node-Specific Recommendations**: Intelligent resource locator suggestions
- Confidence scoring system (0.0 to 1.0) for field-specific recommendations
- High confidence (≥0.8): Exact field matches for known nodes (GitHub owner/repository, Slack channels)
- Medium confidence (≥0.5): Field pattern matches (fields ending in Id, Key, Name)
- Factors: exact field match, field patterns, value patterns, node category
- **Resource Locator Format Detection**: Identifies fields needing `__rl` structure
- Validates resource locator mode (id, url, expression, name, list)
- Auto-fixes missing prefixes in resource locator values
- Provides clear JSON examples showing correct format
- **Enhanced Safety Features**:
- Recursion depth protection (MAX_RECURSION_DEPTH = 100) prevents infinite loops
- Pattern matching precision using exact/prefix matching instead of includes()
- Circular reference detection with WeakSet
- **Separation of Concerns**: Clean architecture for maintainability
- Universal rules separated from node-specific intelligence
- Confidence-based application of suggestions
- Future-proof design that works with any n8n node
## [2.12.1] - 2025-09-22
### Fixed
- **Error Output Validation**: Enhanced workflow validator to detect incorrect error output configurations
- Detects when multiple nodes are incorrectly placed in the same output array (main[0])
- Validates that error handlers are properly connected to main[1] (error output) instead of main[0]
- Cross-validates onError property ('continueErrorOutput') matches actual connection structure
- Provides clear, actionable error messages with JSON examples showing correct configuration
- Uses heuristic detection for error handler nodes (names containing "error", "fail", "catch", etc.)
- Added comprehensive test coverage with 16+ test cases
### Improved
- **Validation Messages**: Error messages now include detailed JSON examples showing both incorrect and correct configurations
- **Pattern Detection**: Fixed `checkWorkflowPatterns` to check main[1] for error outputs instead of non-existent outputs.error
- **Test Coverage**: Added new test file `workflow-validator-error-outputs.test.ts` with extensive error output validation scenarios
## [2.12.0] - 2025-09-19
### Added
- **Flexible Instance Configuration**: Complete multi-instance support for serving multiple n8n instances dynamically
- New `InstanceContext` interface for runtime configuration without multi-tenancy implications
- Dual-mode API client supporting both singleton (env vars) and instance-specific configurations
- LRU cache with SHA-256 hashing for secure client management (100 instances, 30-min TTL)
- Comprehensive input validation preventing injection attacks and invalid configurations
- Session context management in HTTP server for per-session instance configuration
- 100% backward compatibility - existing deployments work unchanged
- Full test coverage with 83 new tests covering security, caching, and validation
### Security
- **SHA-256 Cache Key Hashing**: All instance identifiers are hashed before caching
- **Input Validation**: Comprehensive validation for URLs, API keys, and numeric parameters
- **Secure Logging**: Sensitive data never logged, only partial hashes for debugging
- **Memory Management**: LRU eviction and TTL prevent unbounded growth
- **URL Validation**: Blocks dangerous protocols (file://, javascript://, etc.)
### Performance
- **Efficient Caching**: LRU cache with automatic cleanup reduces API client creation
- **Fast Lookups**: SHA-256 hashed keys for O(1) cache access
- **Memory Optimized**: Maximum 100 concurrent instances with 30-minute TTL
- **Token Savings**: Reuses existing clients instead of recreating
### Documentation
- Added comprehensive [Flexible Instance Configuration Guide](./FLEXIBLE_INSTANCE_CONFIGURATION.md)
- Detailed architecture, usage examples, and security considerations
- Migration guide for existing deployments
- Complete API documentation for InstanceContext
## [2.11.3] - 2025-09-17
### Fixed
- **n8n_update_partial_workflow Tool**: Fixed critical bug where updateNode and updateConnection operations were using incorrect property name
- Changed from `changes` property to `updates` property to match documentation and expected behavior
- Resolves issue where AI agents would break workflow connections when updating nodes
- Fixes GitHub issues #159 (update_partial_workflow is invalid) and #168 (partial workflow update returns error)
- All related tests updated to use correct property name
## [2.11.2] - 2025-09-16
### Updated
- **n8n Dependencies**: Updated to latest versions for compatibility and new features
- n8n: 1.110.1 → 1.111.0
- n8n-core: 1.109.0 → 1.110.0
- n8n-workflow: 1.107.0 → 1.108.0
- @n8n/n8n-nodes-langchain: 1.109.1 → 1.110.0
- **Node Database**: Rebuilt with 535 nodes from updated n8n packages
- **Templates**: Preserved all 2,598 workflow templates with metadata intact
- All critical nodes validated successfully (httpRequest, code, slack, agent)
- Test suite: 1,911 tests passing, 5 flaky performance tests failing (99.7% pass rate)
## [2.11.1] - 2025-09-15
### Added
- **Optional Fields Parameter for search_templates**: Enhanced search_templates tool with field filtering capability
- New optional `fields` parameter accepts an array of field names to include in response
- Supported fields: 'id', 'name', 'description', 'author', 'nodes', 'views', 'created', 'url', 'metadata'
- Reduces response size by 70-98% when requesting only specific fields (e.g., just id and name)
- Maintains full backward compatibility - existing calls without fields parameter work unchanged
- Example: `search_templates({query: "slack", fields: ["id", "name"]})` returns minimal data
- Significantly improves AI agent performance by reducing token usage
### Added
- **Fuzzy Node Type Matching for Templates**: Improved template discovery with flexible node type resolution
- Templates can now be found using simple node names: `["slack"]` instead of `["n8n-nodes-base.slack"]`
- Accepts various input formats: bare names, partial prefixes, and case variations
- Automatically expands related node types: `["email"]` finds Gmail, email send, and related templates
- `["slack"]` also finds `slackTrigger` templates
- Case-insensitive matching: `["Slack"]`, `["WEBHOOK"]`, `["HttpRequest"]` all work
- Backward compatible - existing exact formats continue working
- Reduces failed queries by approximately 50%
- Added `template-node-resolver.ts` utility for node type resolution
- Added 23 tests for template node resolution
- **Structured Template Metadata System**: Comprehensive metadata for intelligent template discovery
- Generated metadata for 2,534 templates (97.5% coverage) using OpenAI's batch API
- Rich metadata structure: categories, complexity, use cases, setup time, required services, key features, target audience
- New `search_templates_by_metadata` tool for advanced filtering by multiple criteria
- Enhanced `list_templates` tool with optional `includeMetadata` parameter
- Templates now always include descriptions in list responses
- Metadata enables filtering by complexity level (simple/medium/complex)
- Filter by estimated setup time ranges (5-480 minutes)
- Filter by required external services (OpenAI, Slack, Google, etc.)
- Filter by target audience (developers, marketers, analysts, etc.)
- Multiple filter combinations supported for precise template discovery
- SQLite JSON extraction for efficient metadata queries
- Batch processing with OpenAI's gpt-4o-mini model for cost efficiency
- Added comprehensive tool documentation for new metadata features
- New database columns: metadata_json, metadata_generated_at
- Repository methods for metadata search and filtering
## [2.11.0] - 2025-01-14
### Added
- **Comprehensive Template Pagination**: All template search and list tools now return paginated responses
- Consistent `PaginatedResponse` format with `items`, `total`, `limit`, `offset`, and `hasMore` fields
- Customizable limits (1-100) and offset parameters for all template tools
- Count methods for accurate pagination information across all template queries
- **New `list_templates` Tool**: Efficient browsing of all available templates
- Returns minimal data (id, name, views, nodeCount) for quick overview
- Supports sorting by views, created_at, or name
- Optimized for discovering templates without downloading full workflow data
- **Flexible Template Retrieval Modes**: Enhanced `get_template` with three response modes
- `nodes_only`: Returns just node types and names (minimal tokens)
- `structure`: Returns nodes with positions and connections (moderate detail)
- `full`: Returns complete workflow JSON (default, maximum detail)
- Reduces token usage by 80-90% in minimal modes
### Enhanced
- **Template Database Compression**: Implemented gzip compression for workflow JSONs
- Workflow data compressed from ~75MB to 12.10MB (84% reduction)
- Database size reduced from 117MB to 48MB despite 5x more templates
- Transparent compression/decompression with base64 encoding
- No API changes - compression is handled internally
- **Template Quality Filtering**: Automatic filtering of low-quality templates
- Templates with ≤10 views are excluded from the database
- Expanded coverage from 499 to 2,596 high-quality templates (5x increase)
- Filtered 4,505 raw templates down to 2,596 based on popularity
- Ensures AI agents work with proven, valuable workflows
- **Enhanced Database Statistics**: Template metrics now included
- Shows total template count, average/min/max views
- Provides complete database overview including template coverage
### Performance
- **Database Optimization**: 59% size reduction while storing 5x more content
- Previous: ~40MB database with 499 templates
- Current: ~48MB database with 2,596 templates
- Without compression would be ~120MB+
- **Token Efficiency**: 80-90% reduction in response size for minimal queries
- `list_templates`: ~10 tokens per template vs 100+ for full data
- `get_template` with `nodes_only`: Returns just essential node information
- Pagination prevents overwhelming responses for large result sets
### Fixed
- **Test Suite Compatibility**: Updated all tests for new template system
- Fixed parameter validation tests to expect new method signatures
- Updated integration tests to use templates with >10 views
- Removed redundant test files that were testing at wrong abstraction level
- All 1,700+ tests now passing
## [2.10.9] - 2025-01-09
### Changed
- **Dependencies**: Updated n8n packages to 1.110.1
- n8n: 1.109.2 → 1.110.1
- n8n-core: 1.108.0 → 1.109.0
- n8n-workflow: 1.106.0 → 1.107.0
- @n8n/n8n-nodes-langchain: 1.108.1 → 1.109.1
### Updated
- **Node Database**: Rebuilt with 536 nodes from updated n8n packages
- **Templates**: Refreshed workflow templates database with latest 499 templates from n8n.io
## [2.10.8] - 2025-09-04
### Updated
- **n8n Dependencies**: Updated to latest versions for compatibility and new features
- n8n: 1.107.4 → 1.109.2
- @n8n/n8n-nodes-langchain: 1.106.2 → 1.109.1
- n8n-nodes-base: 1.106.3 → 1.108.0 (via dependencies)
- **Node Database**: Rebuilt with 535 nodes from updated n8n packages
- **Node.js Compatibility**: Optimized for Node.js v22.17.0 LTS
- Enhanced better-sqlite3 native binary compatibility
- Fixed SQL.js fallback mode for environments without native binaries
- **CI/CD Improvements**: Fixed Rollup native module compatibility for GitHub Actions
- Added explicit platform-specific rollup binaries for cross-platform builds
- Resolved npm ci failures in Linux CI environment
- Fixed package-lock.json synchronization issues
- **Platform Support**: Enhanced cross-platform deployment compatibility
- macOS ARM64 and Linux x64 platform binaries included
- Improved npm package distribution with proper dependency resolution
- All 1,728+ tests passing with updated dependencies
### Fixed
- **CI/CD Pipeline**: Resolved test failures in GitHub Actions
- Fixed pyodide version conflicts between langchain dependencies
- Regenerated package-lock.json with proper dependency resolution
- Fixed Rollup native module loading in Linux CI environment
- **Database Compatibility**: Enhanced SQL.js fallback reliability
- Improved parameter binding and state management
- Fixed statement cleanup to prevent memory leaks
- **Deployment Reliability**: Better handling of platform-specific dependencies
- npm ci now works consistently across development and CI environments
## [2.10.5] - 2025-08-20
### Updated
- **n8n Dependencies**: Updated to latest versions for compatibility and new features
- n8n: 1.106.3 → 1.107.4
- n8n-core: 1.105.3 → 1.106.2
- n8n-workflow: 1.103.3 → 1.104.1
- @n8n/n8n-nodes-langchain: 1.105.3 → 1.106.2
- **Node Database**: Rebuilt with 535 nodes from updated n8n packages
- All tests passing with updated dependencies
## [2.10.4] - 2025-08-12
### Updated
- **n8n Dependencies**: Updated to latest versions for compatibility and new features
- n8n: 1.105.2 → 1.106.3
- n8n-core: 1.104.1 → 1.105.3
- n8n-workflow: 1.102.1 → 1.103.3
- @n8n/n8n-nodes-langchain: 1.104.1 → 1.105.3
- **Node Database**: Rebuilt with 535 nodes from updated n8n packages
- All 1,728 tests passing with updated dependencies
## [2.10.3] - 2025-08-07
### Fixed
- **Validation System Robustness**: Fixed multiple critical validation issues affecting AI agents and workflow validation (fixes #58, #68, #70, #73)
- **Issue #73**: Fixed `validate_node_minimal` crash when config is undefined
- Added safe property access with optional chaining (`config?.resource`)
- Tool now handles undefined, null, and malformed configs gracefully
- **Issue #58**: Fixed `validate_node_operation` crash on invalid nodeType
- Added type checking before calling string methods
- Prevents "Cannot read properties of undefined (reading 'replace')" error
- **Issue #70**: Fixed validation profile settings being ignored
- Extended profile parameter to all validation phases (nodes, connections, expressions)
- Added Sticky Notes filtering to reduce false positives
- Enhanced cycle detection to allow legitimate loops (SplitInBatches)
- **Issue #68**: Added error recovery suggestions for AI agents
- New `addErrorRecoverySuggestions()` method provides actionable recovery steps
- Categorizes errors and suggests specific fixes for each type
- Helps AI agents self-correct when validation fails
### Added
- **Input Validation System**: Comprehensive validation for all MCP tool inputs
- Created `validation-schemas.ts` with custom validation utilities
- No external dependencies - pure TypeScript implementation
- Tool-specific validation schemas for all MCP tools
- Clear error messages with field-level details
- **Enhanced Cycle Detection**: Improved detection of legitimate loops vs actual cycles
- Recognizes SplitInBatches loop patterns as valid
- Reduces false positive cycle warnings
- **Comprehensive Test Suite**: Added 16 tests covering all validation fixes
- Tests for crash prevention with malformed inputs
- Tests for profile behavior across validation phases
- Tests for error recovery suggestions
- Tests for legitimate loop patterns
### Enhanced
- **Validation Profiles**: Now consistently applied across all validation phases
- `minimal`: Reduces warnings for basic validation
- `runtime`: Standard validation for production workflows
- `ai-friendly`: Optimized for AI agent workflow creation
- `strict`: Maximum validation for critical workflows
- **Error Messages**: More helpful and actionable for both humans and AI agents
- Specific recovery suggestions for common errors
- Clear guidance on fixing validation issues
- Examples of correct configurations
## [2.10.2] - 2025-08-05
### Updated
- **n8n Dependencies**: Updated to latest versions for compatibility and new features
- n8n: 1.104.1 → 1.105.2
- n8n-core: 1.103.1 → 1.104.1
- n8n-workflow: 1.101.0 → 1.102.1
- @n8n/n8n-nodes-langchain: 1.103.1 → 1.104.1
- **Node Database**: Rebuilt with 534 nodes from updated n8n packages
- **Template Library**: Fetched 499 workflow templates from the last 12 months
- Templates are filtered to include only those created or updated within the past year
- This ensures the template library contains fresh and actively maintained workflows
- All 1,620 tests passing with updated dependencies
## [2.10.1] - 2025-08-02
### Fixed
- **Memory Leak in SimpleCache**: Fixed critical memory leak causing MCP server connection loss after several hours (fixes #118)
- Added proper timer cleanup in `SimpleCache.destroy()` method
- Updated MCP server shutdown to clean up cache timers
- Enhanced HTTP server error handling with transport error handlers
- Fixed event listener cleanup to prevent accumulation
- Added comprehensive test coverage for memory leak prevention
## [2.10.0] - 2025-08-02
### Added
- **Automated Release System**: Complete CI/CD pipeline for automated releases on version bump
- GitHub Actions workflow (`.github/workflows/release.yml`) with 7 coordinated jobs
- Automatic version detection and changelog extraction
- Multi-artifact publishing: GitHub releases, NPM package, Docker images
- Interactive release preparation tool (`npm run prepare:release`)
- Comprehensive release testing tool (`npm run test:release-automation`)
- Full documentation in `docs/AUTOMATED_RELEASES.md`
- Zero-touch releases: version bump → automatic everything
### Security
- **CI/CD Security Enhancements**:
- Replaced deprecated `actions/create-release@v1` with secure `gh` CLI
- Fixed git checkout vulnerability using safe `git show` commands
- Fixed command injection risk using proper argument arrays
- Added concurrency control to prevent simultaneous releases
- Added disk space checks before resource-intensive operations
- Implemented confirmation gates for destructive operations
### Changed
- **Dockerfile Consolidation**: Removed redundant `Dockerfile.n8n` in favor of single optimized `Dockerfile`
- n8n packages are not required at runtime for N8N_MODE functionality
- Standard image works perfectly with `N8N_MODE=true` environment variable
- Reduces build complexity and maintenance overhead
- Image size reduced by 500MB+ (no unnecessary n8n packages)
- Build time improved from 8+ minutes to 1-2 minutes
### Added (CI/CD Features)
- **Developer Tools**:
- `scripts/prepare-release.js`: Interactive guided release tool
- `scripts/test-release-automation.js`: Validates entire release setup
- `scripts/extract-changelog.js`: Modular changelog extraction
- **Release Automation Features**:
- NPM publishing with 3-retry mechanism for network resilience
- Multi-platform Docker builds (amd64, arm64)
- Semantic version validation and prerelease detection
- Automatic documentation badge updates
- Runtime-optimized NPM package (8 deps vs 50+, ~50MB vs 1GB+)
### Fixed
- Fixed missing `axios` dependency in `package.runtime.json` causing Docker build failures
## [2.9.1] - 2025-08-02
### Fixed
- **Fixed Collection Validation**: Fixed critical issue where AI agents created invalid fixedCollection structures causing "propertyValues[itemName] is not iterable" error (fixes #90)
- Created generic `FixedCollectionValidator` utility class that handles 12 different node types
- Validates and auto-fixes common AI-generated patterns for Switch, If, Filter nodes
- Extended support to Summarize, Compare Datasets, Sort, Aggregate, Set, HTML, HTTP Request, and Airtable nodes
- Added comprehensive test coverage with 19 tests for all affected node types
- Provides clear error messages and automatic structure corrections
- **TypeScript Type Safety**: Improved type safety in fixed collection validator
- Replaced all `any` types with proper TypeScript types (`NodeConfig`, `NodeConfigValue`)
- Added type guards for safe property access
- Fixed potential memory leak in `getAllPatterns` by creating deep copies
- Added circular reference protection using `WeakSet` in structure traversal
- **Node Type Normalization**: Fixed inconsistent node type casing
- Normalized `compareDatasets` to `comparedatasets` and `httpRequest` to `httprequest`
- Ensures consistent node type handling across all validation tools
- Maintains backward compatibility with existing workflows
### Enhanced
- **Code Review Improvements**: Addressed all code review feedback
- Made output keys deterministic by removing `Math.random()` usage
- Improved error handling with comprehensive null/undefined/array checks
- Enhanced memory safety with proper object cloning
- Added protection against circular references in configuration objects
### Testing
- **Comprehensive Test Coverage**: Added extensive tests for fixedCollection validation
- 19 tests covering all 12 affected node types
- Tests for edge cases including empty configs, non-object values, and circular references
- Real-world AI agent pattern tests based on actual ChatGPT/Claude generated configs
- Version compatibility tests across all validation profiles
- TypeScript compilation tests ensuring type safety
## [2.9.0] - 2025-08-01
### Added
- **n8n Integration with MCP Client Tool Support**: Complete n8n integration enabling n8n-mcp to run as MCP server within n8n workflows
- Full compatibility with n8n's MCP Client Tool node
- Dedicated n8n mode (`N8N_MODE=true`) for optimized operation
- Workflow examples and n8n-friendly tool descriptions
- Quick deployment script (`deploy/quick-deploy-n8n.sh`) for easy setup
- Docker configuration specifically for n8n deployment (`Dockerfile.n8n`, `docker-compose.n8n.yml`)
- Test scripts for n8n integration (`test-n8n-integration.sh`, `test-n8n-mode.sh`)
- **n8n Deployment Documentation**: Comprehensive guide for deploying n8n-MCP with n8n (`docs/N8N_DEPLOYMENT.md`)
- Local testing instructions using `/scripts/test-n8n-mode.sh`
- Production deployment with Docker Compose
- Cloud deployment guide for Hetzner, AWS, and other providers
- n8n MCP Client Tool setup and configuration
- Troubleshooting section with common issues and solutions
- **Protocol Version Negotiation**: Intelligent client detection for n8n compatibility
- Automatically detects n8n clients and uses protocol version 2024-11-05
- Standard MCP clients get the latest version (2025-03-26)
- Improves compatibility with n8n's MCP Client Tool node
- Comprehensive protocol negotiation test suite
- **Comprehensive Parameter Validation**: Enhanced validation for all MCP tools
- Clear, user-friendly error messages for invalid parameters
- Numeric parameter conversion and edge case handling
- 52 new parameter validation tests
- Consistent error format across all tools
- **Session Management**: Improved session handling with comprehensive test coverage
- Fixed memory leak potential with async cleanup
- Better connection close handling
- Enhanced session management tests
- **Dynamic README Version Badge**: Made version badge update automatically from package.json
- Added `update-readme-version.js` script
- Enhanced `sync-runtime-version.js` to update README badges
- Version badge now stays in sync during publish workflow
### Fixed
- **Docker Build Optimization**: Fixed Dockerfile.n8n using wrong dependencies
- Now uses `package.runtime.json` instead of full `package.json`
- Reduces build time from 13+ minutes to 1-2 minutes
- Fixes ARM64 build failures due to network timeouts
- Reduces image size from ~1.5GB to ~280MB
- **CI Test Failures**: Resolved Docker entrypoint permission issues
- Updated tests to accept dynamic UID range (10000-59999)
- Enhanced lock file creation with better error recovery
- Fixed TypeScript lint errors in test files
- Fixed flaky performance tests with deterministic versions
- **Schema Validation Issues**: Fixed n8n nested output format compatibility
- Added validation for n8n's nested output workaround
- Fixed schema validation errors with n8n MCP Client Tool
- Enhanced error sanitization for production environments
### Changed
- **Memory Management**: Improved session cleanup to prevent memory leaks
- **Error Handling**: Enhanced error sanitization for production environments
- **Docker Security**: Using unpredictable UIDs/GIDs (10000-59999 range) for better security
- **CI/CD Configuration**: Made codecov patch coverage informational to prevent CI failures on infrastructure code
- **Test Scripts**: Enhanced with Docker auto-installation and better user experience
- Added colored output and progress indicators
- Automatic Docker installation for multiple operating systems
- n8n API key flow for management tools
### Security
- **Enhanced Docker Security**: Dynamic UID/GID generation for containers
- **Error Sanitization**: Improved error messages to prevent information leakage
- **Permission Handling**: Better permission management for mounted volumes
- **Input Validation**: Comprehensive parameter validation prevents injection attacks
## [2.8.3] - 2025-07-31
### Fixed
- **Docker User Switching**: Fixed critical issue where user switching was completely broken in Alpine Linux containers
- Added `su-exec` package for proper privilege dropping in Alpine containers
- Fixed broken shell command in entrypoint that used invalid `exec $*` syntax
- Fixed non-existent `printf %q` command in Alpine's BusyBox shell
- Rewrote user switching logic to properly exec processes with nodejs user
- Fixed race condition in database initialization by ensuring lock directory exists
- **Docker Integration Tests**: Fixed failing tests due to Alpine Linux ps command behavior
- Alpine's BusyBox ps shows numeric UIDs instead of usernames for non-system users
- Tests now accept multiple possible values: "nodejs", "1001", or "1" (truncated)
- Added proper process user verification instead of relying on docker exec output
- Added demonstration test showing docker exec vs main process user context
### Security
- **Command Injection Prevention**: Added comprehensive input validation in n8n-mcp wrapper
- Whitelist-based argument validation to prevent command injection
- Only allows safe arguments: --port, --host, --verbose, --quiet, --help, --version
- Rejects any arguments containing shell metacharacters or suspicious content
- **Database Initialization**: Added proper file locking to prevent race conditions
- Uses flock for exclusive database initialization
- Prevents multiple containers from corrupting database during simultaneous startup
### Testing
- **Docker Test Reliability**: Comprehensive fixes for CI environment compatibility
- Added Docker image build step in test setup
- Fixed environment variable visibility tests to check actual process environment
- Fixed user switching tests to check real process user instead of docker exec context
- All 18 Docker integration tests now pass reliably in CI
### Changed
- **Docker Base Image**: Updated su-exec installation in Dockerfile for proper user switching
- **Error Handling**: Improved error messages and logging in Docker entrypoint script
## [2.8.2] - 2025-07-31
### Added
- **Docker Configuration File Support**: Full support for JSON config files in Docker containers (fixes #105)
- Parse JSON configuration files and safely export as environment variables
- Support for `/app/config.json` mounting in Docker containers
- Secure shell quoting to prevent command injection vulnerabilities
- Dangerous environment variable blocking (PATH, LD_PRELOAD, etc.)
- Key sanitization for invalid environment variable names
- Support for all JSON data types with proper edge case handling
### Fixed
- **Docker Server Mode**: Fixed Docker image failing to start in server mode
- Added `n8n-mcp serve` command support in Docker entrypoint
- Properly set HTTP mode when `serve` command is used
- Fixed missing n8n-mcp binary in Docker image
### Security
- **Command Injection Prevention**: Comprehensive security hardening for config parsing
- Implemented POSIX-compliant shell quoting without using eval
- Blocked dangerous environment variables that could affect system security
- Added protection against shell metacharacters in configuration values
- Sanitized configuration keys to prevent invalid shell variable names
### Testing
- **Docker Configuration Tests**: Added 53 comprehensive tests for Docker config support
- Unit tests for config parsing, security, and edge cases
- Integration tests for Docker entrypoint behavior
- Tests for serve command transformation
- Security-focused tests for injection prevention
### Documentation
- Updated Docker documentation with config file mounting examples
- Added troubleshooting guide for Docker configuration issues
## [2.8.0] - 2025-07-30
### Added
- **Enhanced Test Suite**: Expanded test coverage from 1,182 to 1,356 tests
- **Unit Tests**: Increased from 933 to 1,107 tests across 44 files (was 30)
- Added comprehensive edge case testing for all validators
- Split large test files for better organization and maintainability
- Added test documentation for common patterns and edge cases
- Improved test factory patterns for better test data generation
### Fixed
- **All Test Failures**: Achieved 100% test pass rate (was 99.5%)
- Fixed logger tests by properly setting DEBUG environment variable
- Fixed MSW configuration tests with proper environment restoration
- Fixed workflow validator tests by adding proper connections between nodes
- Fixed TypeScript compilation errors with explicit type annotations
- Fixed ValidationResult mocks to include all required properties
- Fixed environment variable handling in tests for better isolation
### Enhanced
- **Test Organization**: Restructured test files for better maintainability
- Split config-validator tests into 4 focused files: basic, edge-cases, node-specific, security
- Added dedicated edge case test files for all validators
- Improved test naming convention to "should X when Y" pattern
- Better test isolation with proper setup/teardown
### Documentation
- **Test Documentation**: Added comprehensive test guides
- Created test documentation files for common patterns
- Updated test counts in README.md to reflect new test suite
- Added edge case testing guidelines
### CI/CD
- **GitHub Actions**: Fixed permission issues
- Added proper permissions for test, benchmark-pr, and publish workflows
- Fixed status write permissions for benchmark comparisons
- Note: Full permissions will take effect after merge to main branch
## [2.7.23] - 2025-07-30
### Added
- **Comprehensive Testing Infrastructure**: Implemented complete test suite with 1,182 tests
- **933 Unit Tests** across 30 files covering all services, parsers, database, and MCP layers
- **249 Integration Tests** across 14 files for MCP protocol, database operations, and error handling
- **Test Framework**: Vitest with TypeScript, coverage reporting, parallel execution
- **Mock Strategy**: MSW for API mocking, database mocks, MCP SDK test utilities
- **CI/CD**: GitHub Actions workflow with automated testing on all PRs
- **Test Coverage**: Infrastructure in place with lcov, html, and Codecov integration
- **Performance Testing**: Environment-aware thresholds (CI vs local)
- **Database Isolation**: Each test gets its own database for parallel execution
### Fixed
- **CI Test Failures**: Resolved all 115 initially failing integration tests
- Fixed MCP response structure: `response.content[0].text` not `response[0].text`
- Fixed `process.exit(0)` in test setup causing Vitest failures
- Fixed database isolation issues for parallel test execution
- Fixed environment-aware performance thresholds
- Fixed MSW setup isolation preventing interference with unit tests
- Fixed empty database handling in CI environment
- Fixed TypeScript lint errors and strict mode compliance
### Enhanced
- **Test Architecture**: Complete rewrite for production readiness
- Proper test isolation with no shared state
- Comprehensive custom assertions for MCP responses
- Test data generators and builders for complex scenarios
- Environment configuration for test modes
- VSCode integration for debugging
- Meaningful test organization with AAA pattern
### Documentation
- **Testing Documentation**: Complete overhaul to reflect actual implementation
- `docs/testing-architecture.md`: Comprehensive testing guide with real examples
- Documented all 1,182 tests with distribution by component
- Added lessons learned and common issues/solutions
- Updated README with accurate test statistics and badges
### Maintenance
- **Cleanup**: Removed 53 development artifacts and test coordination files
- Deleted temporary agent briefings and coordination documents
- Updated .gitignore to prevent future accumulation
- Cleaned up all `FIX_*.md` and `AGENT_*.md` files
## [2.7.22] - 2025-07-28
### Security
- **Docker base images**: Updated from Node.js 20 Alpine to Node.js 22 LTS Alpine
- Addresses known vulnerabilities in older Alpine images
- Provides better long-term support with Node.js 22 LTS (supported until April 2027)
- All Dockerfiles updated: `Dockerfile`, `Dockerfile.railway`, `Dockerfile.test`
- Docker Compose extractor service updated to use Node.js 22
- Documentation updated to reflect new base image version
### Compatibility
- Tested and verified compatibility with Node.js 22 LTS
- All dependencies work correctly with the new Node.js version
- Docker builds complete successfully with improved security posture
## [2.7.21] - 2025-07-23
### Updated
- **n8n Dependencies**: Updated to latest versions for compatibility and new features
- n8n: 1.102.4 → 1.103.2
- n8n-core: 1.101.2 → 1.102.1
- n8n-workflow: 1.99.1 → 1.100.0
- @n8n/n8n-nodes-langchain: 1.101.2 → 1.102.1
- **Node Database**: Rebuilt with 532 nodes from updated n8n packages
- All validation tests passing with updated dependencies
## [2.7.20] - 2025-07-18
### Fixed
- **Docker container cleanup on session end** (Issue #66)
- Fixed containers not responding to termination signals when Claude Desktop sessions end
- Added proper SIGTERM/SIGINT signal handlers to stdio-wrapper.ts
- Removed problematic trap commands from docker-entrypoint.sh
- Added STOPSIGNAL directive to Dockerfile for explicit signal handling
- Implemented graceful shutdown in MCP server with database cleanup
- Added stdin close detection for proper cleanup when Claude Desktop closes the pipe
- Containers now properly exit with the `--rm` flag, preventing accumulation
- Recommended using `--init` flag in Docker run command for best signal handling
### Documentation
- Updated README with container lifecycle management best practices
- Added `--init` flag to all Docker configuration examples
- Added troubleshooting section for container accumulation issues
## [2.7.19] - 2025-07-18
### Fixed
- **Enhanced node type format normalization** (Issue #74)
- Fixed issue where `n8n-nodes-langchain.chattrigger` (incorrect format) was not being normalized
- Added support for `n8n-nodes-langchain.*` → `nodes-langchain.*` normalization (without @n8n/ prefix)
- Implemented case-insensitive node name matching (e.g., `chattrigger` → `chatTrigger`)
- Added smart camelCase detection for common patterns (trigger, request, sheets, etc.)
- Fixed `get_node_documentation` tool to use same normalization logic as other tools
- All MCP tools now consistently handle various format variations:
- `nodes-langchain.chatTrigger` (correct format)
- `n8n-nodes-langchain.chatTrigger` (package format)
- `n8n-nodes-langchain.chattrigger` (package + wrong case)
- `nodes-langchain.chattrigger` (wrong case only)
- `@n8n/n8n-nodes-langchain.chatTrigger` (full npm format)
- Updated all 7 node lookup locations to use normalized types for alternatives generation
- Enhanced `getNodeTypeAlternatives()` to normalize all generated alternatives
## [2.7.18] - 2025-07-18
### Fixed
- **Node type prefix normalization for AI agents** (Issue #71)
- AI agents can now use node types directly from n8n workflow exports without manual conversion
- Added automatic normalization: `n8n-nodes-base.httpRequest` → `nodes-base.httpRequest`
- Added automatic normalization: `@n8n/n8n-nodes-langchain.agent` → `nodes-langchain.agent`
- Fixed 9 MCP tools that were failing with full package names:
- `get_node_info`, `get_node_essentials`, `get_node_as_tool_info`
- `search_node_properties`, `validate_node_minimal`, `validate_node_config`
- `get_property_dependencies`, `search_nodes`, `get_node_documentation`
- Maintains backward compatibility - existing short prefixes continue to work
- Created centralized `normalizeNodeType` utility for consistent handling across all tools
- **Health check endpoint** - Fixed incorrect `/health` endpoint usage
- Now correctly uses `/healthz` endpoint which is available on all n8n instances
- Improved error handling with proper fallback to workflow list endpoint
- Fixed axios import for healthz endpoint access
- **n8n_list_workflows pagination clarity** (Issue #54)
- Changed misleading `total` field to `returned` to clarify it's the count of items in current page
- Added `hasMore` boolean flag for clear pagination indication
- Added `_note` field with guidance when more data is available ("More workflows available. Use cursor to get next page.")
- Applied same improvements to `n8n_list_executions` for consistency
- AI agents now correctly understand they need to use pagination instead of assuming limited total workflows
### Added
- **Node type utilities** in `src/utils/node-utils.ts`
- `normalizeNodeType()` - Converts full package names to database format
- `getNodeTypeAlternatives()` - Provides fallback options for edge cases
- `getWorkflowNodeType()` - Constructs proper n8n workflow format from database values
- **workflowNodeType field** in all MCP tool responses that return node information
- AI agents now receive both `nodeType` (internal format) and `workflowNodeType` (n8n format)
- Example: `nodeType: "nodes-base.webhook"`, `workflowNodeType: "n8n-nodes-base.webhook"`
- Prevents confusion where AI agents would search nodes and use wrong format in workflows
- Added to: `search_nodes`, `get_node_info`, `get_node_essentials`, `get_node_as_tool_info`, `validate_node_operation`
- **Version information in health check**
- `n8n_health_check` now returns MCP version and supported n8n version
- Added `mcpVersion`, `supportedN8nVersion`, and `versionNote` fields
- Includes instructions for AI agents to inform users about version compatibility
- Note: n8n API currently doesn't expose instance version, so manual verification is required
### Performance
- **n8n_list_workflows response size optimization**
- Tool now returns only minimal metadata (id, name, active, dates, tags, nodeCount) instead of full workflow structure
- Reduced response size by ~95% - from potentially thousands of tokens per workflow to ~10 tokens
- Eliminated token limit errors when listing workflows with many nodes
- Updated tool description to clarify it returns "minimal metadata only"
- Users should use `n8n_get_workflow` to fetch full workflow details when needed
## [2.7.17] - 2025-07-17
### Fixed
- **Removed faulty auto-generated examples from MCP tools** (Issue #60)
- Removed examples from `get_node_essentials` responses that were misleading AI agents
- Removed examples from `validate_node_operation` when validation errors occur
- Examples were showing incorrect configurations (e.g., Slack showing "channel" property instead of required "select" property)
- Tools now focus on validation errors and fix suggestions instead of potentially incorrect examples
- Preserved helpful format hints in `get_node_for_task` (these show input formats like "#general" or URL examples, not node configurations)
- This change reduces confusion and helps AI agents build correct workflows on the first attempt
### Changed
- Updated tool documentation to reflect removal of auto-generated examples
- `get_node_essentials` now points users to `validate_node_operation` for working configurations
- Enhanced validation error messages to be more helpful without relying on examples
## [2.7.16] - 2025-07-17
### Added
- **Comprehensive MCP tools documentation** (Issue #60)
- Documented 30 previously undocumented MCP tools
- Added complete parameter descriptions, examples, and best practices
- Implemented modular documentation system with per-tool files
- Documentation optimized for AI agent consumption (utilitarian approach)
- Added documentation for all n8n management tools (n8n_*)
- Added documentation for workflow validation tools
- Added documentation for template management tools
- Improved `tools_documentation()` to serve as central documentation hub
### Enhanced
- **Tool documentation system** completely rewritten for AI optimization
- Each tool now has its own documentation module
- Consistent structure: description, parameters, examples, tips, common errors
- AI-friendly formatting with clear sections and examples
- Reduced redundancy while maintaining completeness
## [2.7.15] - 2025-07-15
### Fixed
- **HTTP Server URL Handling**: Fixed hardcoded localhost URLs in HTTP server output (Issue #41, #42)
- Added intelligent URL detection that considers BASE_URL, PUBLIC_URL, and proxy headers
- Server now displays correct public URLs when deployed behind reverse proxies
- Added support for X-Forwarded-Proto and X-Forwarded-Host headers when TRUST_PROXY is enabled
- Fixed port display logic to hide standard ports (80/443) in URLs
- Added new GET endpoints (/, /mcp) for better API discovery
### Security
- **Host Header Injection Prevention**: Added hostname validation to prevent malicious proxy headers
- Only accepts valid hostnames (alphanumeric, dots, hyphens, optional port)
- Rejects hostnames with paths, usernames, or special characters
- Falls back to safe defaults when invalid headers are detected
- **URL Scheme Validation**: Restricted URL schemes to http/https only
- Blocks dangerous schemes like javascript:, file://, data:
- Validates all configured URLs (BASE_URL, PUBLIC_URL)
- **Information Disclosure**: Removed sensitive environment data from API responses
- Root endpoint no longer exposes internal configuration
- Only shows essential API information
### Added
- **URL Detection Utility**: New `url-detector.ts` module for intelligent URL detection
- Prioritizes explicit configuration (BASE_URL/PUBLIC_URL)
- Falls back to proxy headers when TRUST_PROXY is enabled
- Uses host/port configuration as final fallback
- Includes comprehensive security validations
- **Test Scripts**: Added test scripts for URL configuration and security validation
- `test-url-configuration.ts`: Tests various URL detection scenarios
- `test-security.ts`: Validates security fixes for malicious headers
### Changed
- **Consistent Versioning**: Fixed version inconsistency between server implementations
- Both http-server.ts and http-server-single-session.ts now use PROJECT_VERSION
- Removed hardcoded version strings
- **HTTP Bridge**: Updated to use HOST/PORT environment variables for default URL construction
- **Documentation**: Updated HTTP deployment guide with URL configuration section
## [2.7.14] - 2025-07-15
### Fixed
- **Partial Update Tool**: Fixed validation/execution discrepancy that caused "settings must NOT have additional properties" error (Issue #45)
- Removed logic in `cleanWorkflowForUpdate` that was incorrectly adding default settings to workflows
- The function now only removes read-only fields without adding any new properties
- This fixes the issue where partial updates would pass validation but fail during execution
- Added comprehensive test coverage in `test-issue-45-fix.ts`
## [2.7.13] - 2025-07-11
### Fixed
- **npx Execution**: Fixed WASM file resolution for sql.js when running via `npx n8n-mcp` (Issue #31)
- Enhanced WASM file locator to try multiple path resolution strategies
- Added `require.resolve()` for reliable package location in npm environments
- Made better-sqlite3 an optional dependency to prevent installation failures
- Improved error messages when sql.js fails to load
- The package now works correctly with `npx` without any manual configuration
### Changed
- **Database Adapter**: Improved path resolution for both local development and npm package contexts
- Supports various npm installation scenarios (global, local, npx cache)
- Better fallback handling for sql.js WebAssembly file loading
## [2.7.12] - 2025-07-10
### Updated
- **n8n Dependencies**: Updated to latest versions for compatibility and new features
- n8n: 1.100.1 → 1.101.1
- n8n-core: 1.99.0 → 1.100.0
- n8n-workflow: 1.97.0 → 1.98.0
- @n8n/n8n-nodes-langchain: 1.99.0 → 1.100.1
- **Node Database**: Rebuilt with 528 nodes from updated n8n packages
- All validation tests passing with updated dependencies
## [2.7.11] - 2025-07-10
### Enhanced
- **Token Efficiency**: Significantly reduced MCP tool description lengths for better AI agent performance
- Documentation tools: Average 129 chars (down from ~250-450)
- Management tools: Average 93 chars (down from ~200-400)
- Overall token reduction: ~65-70%
- Moved detailed documentation to `tools_documentation()` system
- Only 2 tools exceed 200 chars (list_nodes: 204, n8n_update_partial_workflow: 284)
- Preserved all essential information while removing redundancy
### Fixed
- **search_nodes Tool**: Major improvements to search functionality for AI agents
- Primary nodes (webhook, httpRequest) now appear first in search results instead of being buried
- Fixed issue where searching "webhook" returned specialized triggers instead of the main Webhook node
- Fixed issue where searching "http call" didn't prioritize HTTP Request node
- Fixed FUZZY mode returning no results for typos like "slak" (lowered threshold from 300 to 200)
- Removed unnecessary searchInfo messages that appeared on every search
- Fixed HTTP node type comparison case sensitivity issue
- Implemented relevance-based ranking with special boosting for primary nodes
- **search_templates FTS5 Error**: Fixed "no such module: fts5" error in environments without FTS5 support (fixes Claude Desktop issue)
- Made FTS5 completely optional - detects support at runtime
- Removed FTS5 from required schema to prevent initialization failures
- Automatically falls back to LIKE search when FTS5 is unavailable
- FTS5 tables and triggers created conditionally only if supported
- Template search now works in ALL SQLite environments
### Added
- **FTS5 Full-Text Search**: Added SQLite FTS5 support for faster and more intelligent node searching
- Automatic fallback to LIKE queries if FTS5 is unavailable
- Supports advanced search modes: OR (default), AND (all terms required), FUZZY (typo-tolerant)
- Significantly improves search performance for large databases
- FUZZY mode now uses edit distance (Levenshtein) for better typo tolerance
- **FTS5 Detection**: Added runtime detection of FTS5 support
- `checkFTS5Support()` method in database adapters
- Conditional initialization of FTS5 features
- Graceful degradation when FTS5 not available
## [Unreleased]
### Fixed
- **Code Node Documentation**: Corrected information about `$helpers` object and `getWorkflowStaticData` function
- `$getWorkflowStaticData()` is a standalone function, NOT `$helpers.getWorkflowStaticData()`
- Updated Code node guide to clarify which functions are standalone vs methods on $helpers
- Added validation warning when using incorrect `$helpers.getWorkflowStaticData` syntax
- Based on n8n community feedback and GitHub issues showing this is a common confusion point
### Added
- **Expression vs Code Node Clarification**: Added comprehensive documentation about differences between expression and Code node contexts
- New section "IMPORTANT: Code Node vs Expression Context" explaining key differences
- Lists expression-only functions not available in Code nodes ($now(), $today(), Tournament template functions)
- Clarifies different syntax: $('Node Name') vs $node['Node Name']
- Documents reversed JMESPath parameter order between contexts
- Added "Expression Functions NOT in Code Nodes" section with alternatives
- **Enhanced Code Node Validation**: Added new validation checks for common expression/Code node confusion
- Detects expression syntax {{...}} in Code nodes with clear error message
- Warns about using $node[] syntax instead of $() in Code nodes
- Identifies expression-only functions with helpful alternatives
- Checks for wrong JMESPath parameter order
- Test script `test-expression-code-validation.ts` to verify validation works correctly
## [2.7.11] - 2025-07-09
### Fixed
- **Issue #26**: Fixed critical issue where AI agents were placing error handling properties inside `parameters` instead of at node level
- Root cause: AI agents were confused by examples showing `parameters.path` updates and assumed all properties followed the same pattern
- Error handling properties (`onError`, `retryOnFail`, `maxTries`, `waitBetweenTries`, `alwaysOutputData`) must be placed at the NODE level
- Other node-level properties (`executeOnce`, `disabled`, `notes`, `notesInFlow`, `credentials`) were previously undocumented for AI agents
- Updated `n8n_create_workflow` and `n8n_update_partial_workflow` documentation with explicit examples and warnings
- Verified fix with workflows tGyHrsBNWtaK0inQ, usVP2XRXhI35m3Ts, and swuogdCCmNY7jj71
### Added
- **Comprehensive Node-Level Properties Reference** in tools documentation (`tools_documentation()`)
- Documents ALL available node-level properties with explanations
- Shows correct placement and usage for each property
- Provides complete example node configuration
- Accessible via `tools_documentation({depth: "full"})` for AI agents
- **Enhanced Workflow Validation** for additional node-level properties
- Now validates `executeOnce`, `disabled`, `notes`, `notesInFlow` types
- Checks for misplacement of ALL node-level properties (expanded from 6 to 11)
- Provides clear error messages with correct examples when properties are misplaced
- Shows specific fix with example node structure
- **Test Script** `test-node-level-properties.ts` demonstrating correct usage
- Shows all node-level properties in proper configuration
- Demonstrates common mistakes to avoid
- Validates workflow configurations
- **Comprehensive Code Node Documentation** in tools_documentation
- New `code_node_guide` topic with complete reference for JavaScript and Python
- Covers all built-in variables: $input, $json, $node, $workflow, $execution, $prevNode
- Documents helper functions: DateTime (Luxon), JMESPath, $helpers methods
- Includes return format requirements with correct/incorrect examples
- Security considerations and banned operations
- Common patterns: data transformation, filtering, aggregation, error handling
- Code node as AI tool examples
- Performance best practices and debugging tips
- **Enhanced Code Node Validation** with n8n-specific patterns
- Validates return statement presence and format
- Checks for array of objects with json property
- Detects common mistakes (returning primitives, missing array wrapper)
- Validates n8n variable usage ($input, items, $json context)
- Security checks (eval, exec, require, file system access)
- Language-specific validation for JavaScript and Python
- Mode-specific warnings ($json in wrong mode)
- Async/await pattern validation
- External library detection with helpful alternatives
- **Expanded Code Node Examples** in ExampleGenerator
- Data transformation, aggregation, and filtering examples
- API integration with error handling
- Python data processing example
- Code node as AI tool pattern
- CSV to JSON transformation
- All examples include proper return format
- **New Code Node Task Templates**
- `custom_ai_tool`: Create custom tools for AI agents
- `aggregate_data`: Summary statistics from multiple items
- `batch_process_with_api`: Process items in batches with rate limiting
- `error_safe_transform`: Robust data transformation with validation
- `async_data_processing`: Concurrent processing with limits
- `python_data_analysis`: Statistical analysis using Python
- All templates include comprehensive error handling
- **Fixed Misleading Documentation** based on real-world testing:
- **Crypto Module**: Clarified that `require('crypto')` IS available despite editor warnings
- **Helper Functions**: Fixed documentation showing `$getWorkflowStaticData()` is standalone, not on $helpers
- **JMESPath**: Corrected syntax from `jmespath.search()` to `$jmespath()`
- **Node Access**: Fixed from `$node['Node Name']` to `$('Node Name')`
- **Python**: Documented `item.json.to_py()` for JsProxy conversion
- Added comprehensive "Available Functions and Libraries" section
- Created security examples showing proper crypto usage
- **JMESPath Numeric Literals**: Added critical documentation about n8n-specific requirement for backticks around numbers in filters
- Example: `[?age >= \`18\`]` not `[?age >= 18]`
- Added validation to detect and warn about missing backticks
- Based on Claude Desktop feedback from workflow testing
- **Webhook Data Structure**: Fixed common webhook data access gotcha
- Webhook payload is at `items[0].json.body`, NOT `items[0].json`
- Added dedicated "Webhook Data Access" section in Code node documentation
- Created webhook processing example showing correct data access
- Added validation to detect incorrect webhook data access patterns
- New task template `process_webhook_data` with complete example
### Enhanced
- **MCP Tool Documentation** significantly improved:
- `n8n_create_workflow` now includes complete node example with all properties
- `n8n_update_partial_workflow` shows difference between node-level vs parameter updates
- Added "CRITICAL" warnings about property placement
- Updated best practices and common pitfalls sections
- **Workflow Validator** improvements:
- Expanded property checking from 6 to 11 node-level properties
- Better error messages showing complete correct structure
- Type validation for all node-level boolean and string properties
- **Code Node Validation** enhanced with new checks:
- Detects incorrect `$helpers.getWorkflowStaticData()` usage
- Warns about `$helpers` usage without availability check
- Validates crypto usage with proper require statement
- All based on common errors found in production workflows
- **Type Definitions** updated:
- Added `notesInFlow` to WorkflowNode interface in workflow-validator.ts
- Fixed credentials type from `Record<string, string>` to `Record<string, unknown>` in n8n-api.ts
- **NodeSpecificValidators** now includes comprehensive Code node validation
- Language-specific syntax checks
- Return format validation with detailed error messages
- n8n variable usage validation
- Security pattern detection
- Error handling recommendations
- Mode-specific suggestions
- **Config Validator** improved Code node validation
- Better return statement detection
- Enhanced syntax checking for both JavaScript and Python
- More helpful error messages with examples
- Detection of common n8n Code node mistakes
- **Fixed Documentation Inaccuracies** based on user testing and n8n official docs:
- JMESPath: Corrected syntax to `$jmespath()` instead of `jmespath.search()`
- Node Access: Fixed to show `$('Node Name')` syntax, not `$node`
- Python: Documented `_input.all()` and `item.json.to_py()` for JsProxy conversion
- Python: Added underscore prefix documentation for all built-in variables
- Validation: Skip property visibility warnings for Code nodes to reduce false positives
## [2.7.10] - 2025-07-09
### Documentation Update
- Added comprehensive documentation on how to update error handling properties using `n8n_update_partial_workflow`
- Error handling properties can be updated at the node level using the workflow diff engine:
- `continueOnFail`: boolean - Whether to continue workflow on node failure
- `onError`: 'continueRegularOutput' | 'continueErrorOutput' | 'stopWorkflow' - Error handling strategy
- `retryOnFail`: boolean - Whether to retry on failure
- `maxTries`: number - Maximum retry attempts
- `waitBetweenTries`: number - Milliseconds to wait between retries
- `alwaysOutputData`: boolean - Always output data even on error
- Added test script demonstrating error handling property updates
- Updated WorkflowNode type to include `onError` property in n8n-api types
- Workflow diff engine now properly handles all error handling properties
## [2.7.10] - 2025-07-07
### Added
- Enhanced authentication logging for better debugging of client authentication issues
- Specific error reasons for authentication failures: `no_auth_header`, `invalid_auth_format`, `invalid_token`
- AUTH_TOKEN_FILE support in single-session HTTP server for consistency
- Empty token validation to prevent security issues
- Whitespace trimming for authentication tokens
### Fixed
- Issue #22: Improved authentication failure diagnostics for mcp-remote client debugging
- Issue #16: Fixed AUTH_TOKEN_FILE validation for HTTP mode in Docker production stacks - Docker entrypoint now properly validates and supports AUTH_TOKEN_FILE environment variable
- Security: Removed token length from logs to prevent information disclosure
### Security
- Authentication tokens are now trimmed to handle whitespace edge cases
- Empty tokens are explicitly rejected with clear error messages
- Removed sensitive information (token lengths) from authentication logs
## [2.7.8] - 2025-07-06
### Added
- npx support for zero-installation usage - users can now run `npx n8n-mcp` without installing
- npm package distribution with runtime-only dependencies (8 deps vs 50+ dev deps)
- Dedicated publish script for npm releases with OTP support
- Database path resolution supporting npx, global, and local installations
### Fixed
- Issue #15: Added npx execution support as requested
- Removed development dependencies from npm package (reduced from 1GB+ to ~50MB)
- Node.js version conflicts by excluding n8n dependencies from runtime package
### Changed
- npm package now uses package.runtime.json for publishing (no n8n dependencies)
- Enhanced .gitignore to exclude npm publishing artifacts
- README now highlights npx as the primary installation method
## [2.7.5] - 2025-07-06
### Added
- AUTH_TOKEN_FILE support for reading authentication tokens from files (Docker secrets compatible) - partial implementation
- Known Issues section in README documenting Claude Desktop container duplication bug
- Enhanced authentication documentation in Docker README
### Fixed
- Issue #16: AUTH_TOKEN_FILE was documented but not implemented (partially fixed - see v2.7.10 for complete fix)
- HTTP server now properly supports both AUTH_TOKEN and AUTH_TOKEN_FILE environment variables
### Changed
- Authentication logic now checks AUTH_TOKEN first, then falls back to AUTH_TOKEN_FILE
- Updated Docker documentation to clarify authentication options
## [2.7.4] - 2025-07-03
### Changed
- Renamed `start_here_workflow_guide` tool to `tools_documentation` for better clarity
- Converted tool output from JSON to LLM-friendly plain text format
- Made documentation concise by default with "essentials" depth
### Added
- `depth` parameter to control documentation detail level ("essentials" or "full")
- Per-tool documentation - get help for any specific MCP tool
- Two-tier documentation system:
- Essentials: Brief description, key parameters, example, performance, tips
- Full: Complete documentation with all details, examples, best practices
- Quick reference mode when called without parameters
- Documentation for 8 commonly used tools
- Test script for tools documentation (`test:tools-documentation`)
### Removed
- Removed duplicate `tools_documentation` tool definition
- Removed unused `getWorkflowGuide` method (380+ lines)
- Removed old `handlers-documentation.ts` file
## [2.7.3] - 2025-07-02
### Added
- MCP Tools Documentation system (initial implementation)
- `tools_documentation` tool for comprehensive MCP tool documentation
- Documentation includes parameters, examples, best practices, and pitfalls
- Search tools by keyword functionality
- Browse tools by category
- Quick reference guide with workflow patterns
### Fixed
- Cleaned up redundant tool definitions
## [2.7.2] - 2025-07-01
### Fixed
- HTTP deployment documentation improvements
- Docker configuration updates with n8n API options
### Changed
- Updated version handling in multiple configuration files
## [2.7.1] - 2025-06-30
### Fixed
- Workflow diff engine edge cases
- Transactional update processing improvements
### Added
- Additional test coverage for diff operations
- Debug scripts for update operations
## [2.7.0] - 2025-06-29
### Added
- New `n8n_update_partial_workflow` tool for efficient diff-based workflow editing with transactional updates
- WorkflowDiffEngine for applying targeted edits without sending full workflow JSON (80-90% token savings)
- 13 diff operations: addNode, removeNode, updateNode, moveNode, enableNode, disableNode, addConnection, removeConnection, updateConnection, updateSettings, updateName, addTag, removeTag
- Smart node references supporting both node ID and name
- Transaction safety with validation before applying changes
- Validation-only mode for testing diff operations
- Comprehensive test coverage for all diff operations
- Example guide in `docs/workflow-diff-examples.md`
- Two-pass processing allowing operations in any order
- Operation limit of 5 operations per request for reliability
- `n8n_diagnostic` tool to troubleshoot management tools visibility issues
- Version utility (`src/utils/version.ts`) for centralized version management
- Script to sync package.runtime.json version
### Changed
- Renamed `n8n_update_workflow` to `n8n_update_full_workflow` to clarify it replaces entire workflow
- Renamed core MCP files for clarity:
- `tools-update.ts` → `tools.ts`
- `server-update.ts` → `server.ts`
- `http-server-fixed.ts` → `http-server.ts`
- Updated imports across 21+ files to use new file names
### Fixed
- Version mismatch issue where version was hardcoded instead of reading from package.json (GitHub issue #5)
- MCP validation error by simplifying schema to allow additional properties
- n8n API validation by removing all read-only fields in cleanWorkflowForUpdate
- Claude Desktop compatibility by adding additionalProperties: true
- Removed DEBUG console.log statements from MCP server
### Removed
- Legacy HTTP server implementation (`src/http-server.ts`)
- Unused legacy API client (`src/utils/n8n-client.ts`)
- Unnecessary file name suffixes (-update, -fixed)
## [2.6.3] - 2025-06-26
### Added
- `n8n_validate_workflow` tool to validate workflows directly from n8n instance by ID
- Fetches workflow from n8n API and runs comprehensive validation
- Supports all validation profiles and options
- Part of complete lifecycle: discover → build → validate → deploy → execute
## [2.6.2] - 2025-06-26
### Added
- Node type validation to verify node types exist in n8n
- Smart suggestions for common mistakes (e.g., `webhook` → `n8n-nodes-base.webhook`)
- Minimum viable workflow validation preventing single-node workflows (except webhooks)
- Empty connection detection for multi-node workflows
- Helper functions: `getWorkflowStructureExample()` and `getWorkflowFixSuggestions()`
### Fixed
- nodes-base prefix detection now catches errors before database lookup
- Enhanced error messages with clear guidance on proper workflow structure
## [2.6.1] - 2025-06-26
### Added
- typeVersion validation in workflow validator
- Enforces typeVersion on all versioned nodes
- Warns on outdated node versions
- Prevents invalid version numbers
### Fixed
- Missing typeVersion errors with correct version suggestions
- Invalid version detection exceeding maximum supported
## [2.6.0] - 2025-06-26
### Added
- 14 n8n management tools for complete workflow lifecycle management:
- `n8n_create_workflow` - Create workflows programmatically
- `n8n_update_workflow` - Update existing workflows
- `n8n_trigger_webhook_workflow` - Execute workflows via webhooks
- `n8n_list_executions` - Monitor workflow executions
- `n8n_health_check` - Check n8n instance connectivity
- And 9 more workflow and execution management tools
- Integration with n8n-manager-for-ai-agents functionality
- Conditional tool registration based on N8N_API_URL and N8N_API_KEY configuration
- Smart error handling for API limitations
## [2.5.1] - 2025-06-24
### Added
- `get_node_as_tool_info` tool for specific information about using ANY node as an AI tool
- Enhanced AI tool support with usage guidance
- Improved start_here_workflow_guide with Claude Project setup
### Changed
- Enhanced AI tool detection and documentation
- Updated documentation to match current state
## [2.5.0] - 2025-06-24
### Added
- Comprehensive workflow validation system:
- `validate_workflow` - Validate entire workflows before deployment
- `validate_workflow_connections` - Check workflow structure and connections
- `validate_workflow_expressions` - Validate all n8n expressions
- Expression validator for n8n syntax validation
- AI tool connection validation
- Phase 2 validation improvements
## [2.4.2] - 2025-06-24
### Added
- Enhanced operation-aware validation system
- `validate_node_operation` - Verify node configuration with operation awareness
- `validate_node_minimal` - Quick validation for required fields only
- Node-specific validation logic
- Validation profiles support
### Fixed
- Validation improvements based on AI agent feedback
## [2.4.1] - 2025-06-20
### Added
- n8n workflow templates integration:
- `list_node_templates` - Find workflow templates using specific nodes
- `get_template` - Get complete workflow JSON for import
- `search_templates` - Search templates by keywords
- `get_templates_for_task` - Get curated templates for common tasks
- Template fetching from n8n.io API
- Robust template fetching with retries
- Expanded template window from 6 months to 1 year
### Fixed
- Made templates available in Docker by removing axios from runtime
- Template service made optional in Docker environment
- Non-deterministic CHECK constraint removed from templates table
## [2.4.0] - 2025-06-18
### Added
- AI-optimized tools with 95% size reduction:
- `get_node_essentials` - Returns only essential properties (10-20) with examples
- `search_node_properties` - Find specific properties without downloading everything
- `get_node_for_task` - Get pre-configured node settings for common tasks
- `list_tasks` - List all available task templates
- `get_property_dependencies` - Analyze property dependencies and visibility conditions
- Property filter service with curated essential properties
- Example generator for common use cases
- Task templates with pre-configured settings
- Docker build optimization (82% smaller images, 10x faster builds)
### Changed
- Switched to MIT license for wider adoption
- Optimized Docker builds to exclude n8n dependencies at runtime
- Improved tool descriptions and added workflow guide tool
### Fixed
- Docker build failures in GitHub Actions
- Claude Desktop stdio communication issues
- Version array handling in node parser
### Removed
- Legacy MCP implementation files
- n8n dependencies from Docker runtime image
## [2.3.3] - 2025-06-16
### Added
- Smart dependency update system for n8n packages
- GitHub Actions workflow for automated n8n updates
- Alternative Renovate configuration
### Fixed
- n8n package interdependent version requirements
- Node type references in validation script
## [2.3.2] - 2025-06-14
### Added
- Single-session HTTP server architecture
- Direct JSON-RPC implementation for HTTP mode
- Console output isolation for clean JSON-RPC responses
### Fixed
- "stream is not readable" error in HTTP server
- "Server not initialized" error with StreamableHTTPServerTransport
- MCP HTTP server stream errors
## [2.3.1] - 2025-06-13
### Added
- HTTP server mode for remote deployment with token authentication
- MCP-compatible HTTP endpoints
- Security features: CORS, rate limiting, request size limits
- Comprehensive HTTP testing scripts
## [2.3.0] - 2025-06-12
### Added
- Universal Node.js compatibility with automatic database adapter fallback
- Database adapter pattern with BetterSQLiteAdapter and SQLJSAdapter
- Automatic adapter selection based on environment
- sql.js persistence layer with debounced saves
### Changed
- Database operations now use unified adapter interface
- Transparent adapter switching for different Node.js versions
## [2.2.0] - 2025-06-12
### Added
- Enhanced node parser with versioned node support
- Dedicated property extractor for complex node structures
- Full support for @n8n/n8n-nodes-langchain package
- AI tool detection (35 tools with usableAsTool property)
### Changed
- Major refactor based on IMPLEMENTATION_PLAN.md v2.2
- Improved property/operation extraction (452/458 nodes have properties)
- Enhanced documentation mapping
### Fixed
- VersionedNodeType handling
- Documentation mapping issues
- Property extraction for 98.7% of nodes
## [2.1.0] - 2025-06-09
### Added
- Node extraction scripts for n8n modules
- Docker setup for n8n module processing
- Enhanced documentation fetcher
- Node source extractor utility
## [2.0.0] - 2025-06-08
### Added
- Complete overhaul to enhanced documentation-only MCP server
- SQLite database with FTS5 for fast searching
- Comprehensive MCP tools for querying n8n nodes
- Node documentation service as core component
### Changed
- Architecture redesign focusing on documentation serving
- Removed workflow execution capabilities
- Simplified to documentation and knowledge serving
## [1.0.0] - 2025-06-08
### Added
- Initial release
- Basic n8n and MCP integration
- Core workflow automation features
[2.12.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.11.3...v2.12.0
[2.11.3]: https://github.com/czlonkowski/n8n-mcp/compare/v2.11.2...v2.11.3
[2.11.2]: https://github.com/czlonkowski/n8n-mcp/compare/v2.11.1...v2.11.2
[2.11.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.11.0...v2.11.1
[2.11.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.9...v2.11.0
[2.10.9]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.8...v2.10.9
[2.10.8]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.5...v2.10.8
[2.10.5]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.4...v2.10.5
[2.10.4]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.3...v2.10.4
[2.10.3]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.2...v2.10.3
[2.10.2]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.1...v2.10.2
[2.10.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.10.0...v2.10.1
[2.10.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.9.1...v2.10.0
[2.9.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.9.0...v2.9.1
[2.9.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.8.3...v2.9.0
[2.8.3]: https://github.com/czlonkowski/n8n-mcp/compare/v2.8.2...v2.8.3
[2.8.2]: https://github.com/czlonkowski/n8n-mcp/compare/v2.8.0...v2.8.2
[2.8.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.23...v2.8.0
[2.7.23]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.22...v2.7.23
[2.7.22]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.21...v2.7.22
[2.7.21]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.20...v2.7.21
[2.7.20]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.19...v2.7.20
[2.7.19]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.18...v2.7.19
[2.7.18]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.17...v2.7.18
[2.7.17]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.16...v2.7.17
[2.7.16]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.15...v2.7.16
[2.7.15]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.13...v2.7.15
[2.7.13]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.12...v2.7.13
[2.7.12]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.11...v2.7.12
[2.7.11]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.10...v2.7.11
[2.7.10]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.8...v2.7.10
[2.7.8]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.5...v2.7.8
[2.7.5]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.4...v2.7.5
[2.7.4]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.3...v2.7.4
[2.7.3]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.2...v2.7.3
[2.7.2]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.1...v2.7.2
[2.7.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.7.0...v2.7.1
[2.7.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.6.3...v2.7.0
[2.6.3]: https://github.com/czlonkowski/n8n-mcp/compare/v2.6.2...v2.6.3
[2.6.2]: https://github.com/czlonkowski/n8n-mcp/compare/v2.6.1...v2.6.2
[2.6.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.6.0...v2.6.1
[2.6.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.5.1...v2.6.0
[2.5.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.5.0...v2.5.1
[2.5.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.4.2...v2.5.0
[2.4.2]: https://github.com/czlonkowski/n8n-mcp/compare/v2.4.1...v2.4.2
[2.4.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.4.0...v2.4.1
[2.4.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.3.3...v2.4.0
[2.3.3]: https://github.com/czlonkowski/n8n-mcp/compare/v2.3.2...v2.3.3
[2.3.2]: https://github.com/czlonkowski/n8n-mcp/compare/v2.3.1...v2.3.2
[2.3.1]: https://github.com/czlonkowski/n8n-mcp/compare/v2.3.0...v2.3.1
[2.3.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.2.0...v2.3.0
[2.2.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.1.0...v2.2.0
[2.1.0]: https://github.com/czlonkowski/n8n-mcp/compare/v2.0.0...v2.1.0
[2.0.0]: https://github.com/czlonkowski/n8n-mcp/compare/v1.0.0...v2.0.0
[1.0.0]: https://github.com/czlonkowski/n8n-mcp/releases/tag/v1.0.0