v2.1.4

fix: Increase proxy mode timeout from 30s to 4 minutes for AI operations

- Server proxy timeout: 30s → 4 minutes (configurable via PROXY_AI_TIMEOUT_MS)
- Frontend AI timeout: 2 minutes → 4 minutes for all AI operations
- Handles 3+ minute AI responses in proxy mode without timing out
- Added environment variable PROXY_AI_TIMEOUT_MS for deployment flexibility
- Enhanced timeout error messages with duration details
- Updated tests to match new timeout values

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

v2.1.3

🚀 Agentic Markdown Todos v2.1.3

🐛 Critical Bug Fix: Proxy Mode Timeout Issue

Problem Resolved

Fixed a critical timeout issue where proxy mode AI requests were timing out at 30 seconds while AI processing could take up to 3+ minutes, causing request failures in local AI proxy configurations.

Key Fixes

• Server Proxy Timeout: Increased from 30 seconds → 4 minutes (configurable via PROXY_AI_TIMEOUT_MS)
• Frontend AI Timeout: Increased from 2 minutes → 4 minutes for all AI operations
• Environment Variable Support: Added PROXY_AI_TIMEOUT_MS for deployment flexibility
• Enhanced Error Messages: Timeout errors now include duration details for better debugging

Technical Details

# Configure custom timeout (optional)
export PROXY_AI_TIMEOUT_MS=300000  # 5 minutes

Before: Browser → Server → Proxy (30s timeout) ❌ AI Response (3+ minutes)
After: Browser → Server → Proxy (4min timeout) ✅ AI Response (3+ minutes)

Impact

• ✅ Local AI Proxy: Now handles long-running AI operations without timeout
• ✅ Large Model Support: Compatible with slower local AI models
• ✅ Enterprise Deployments: Configurable timeouts for different environments
• ✅ Backward Compatibility: No breaking changes for existing users

🏠 Major Feature: Local AI Proxy Implementation

Revolutionary AI Independence

This release includes the complete Local AI Proxy as a first-class AI provider, enabling complete data sovereignty and AI processing independence:

• 🔐 Complete Data Control: Process ALL AI requests locally without external service dependency
• 💰 Cost Elimination: Avoid external API costs with local AI infrastructure
• 🏢 Enterprise Compliance: Meet strict data governance requirements
• 🛠️ Multi-Platform Support: Compatible with LMStudio, Ollama, and custom OpenAI endpoints

Key Technical Features

• Provider-Based Architecture: Simple, reliable AI request routing
• Real-Time Connection Monitoring: Live status validation with 30-second health checks
• WebSocket Infrastructure: Efficient proxy communication system
• User-Specific Authentication: UUID + localToken security model
• Enhanced Setup UI: Tabbed configuration wizard with comprehensive guidance

Docker Integration

# LMStudio Configuration
docker run -d --name am_todos_proxy \
  -v ./proxy-config:/config \
  -e LOCAL_PROXY_MODE=true \
  -e MAIN_SERVER_TOKEN=<your-token> \
  -e MAIN_SERVER_URL=wss://your-app.run.app/proxy-ws \
  -e LOCAL_AI_ENDPOINT=http://host.docker.internal:1234 \
  -e LOCAL_AI_MODEL=mistralai/mistral-small-3.2 \
  ghcr.io/florinpeter/am-todos:v2.1.3

📊 Quality Assurance

Comprehensive Testing

• 1,017 Tests Passing: Complete test suite validation
• 79.4% Code Coverage: Extensive coverage across all components
• Timeout Fix Validation: All fetchWithTimeout tests updated and passing
• Multi-Platform Testing: Node.js 20.x and 22.x validation

Production Ready

• Enhanced Error Handling: Comprehensive timeout error states with recovery guidance
• Performance Optimized: Proper timeout handling without request race conditions
• Mobile Responsive: Complete mobile-first design with touch-friendly interface

🌟 Business Impact

For Organizations

• Data Sovereignty: Complete control over AI processing and data
• Compliance Ready: Meet strict enterprise security requirements
• Cost Control: Eliminate external AI API expenses
• Infrastructure Independence: Use existing local AI investments
• Configurable Timeouts: Adapt to different deployment scenarios

For Users

• Seamless Experience: Local proxy works identically to cloud providers
• Real-Time Feedback: Live connection status and troubleshooting guidance
• Flexible Setup: Support for popular local AI platforms
• Enhanced Security: User-specific proxy routing with credential validation
• Reliable AI Processing: No more timeout failures on complex AI tasks

🔄 Migration Notes

Automatic Updates

• Timeout Improvements: Automatic benefit from increased timeout values
• Settings Migration: Automatic localStorage updates
• Test Validation: All existing functionality preserved and improved

New Users

1. Select AI Provider: Choose "Local Proxy" alongside Gemini and OpenRouter
2. Deploy Proxy: Use provided Docker commands for your AI platform
3. Configure Credentials: Copy UUID and localToken from proxy settings
4. Test Connection: Validate setup with real-time status monitoring
5. Start Using: All AI features work seamlessly with local processing

---

🏆 Complete Success: This release fixes critical timeout issues while maintaining all the revolutionary Local AI Proxy features, providing unprecedented control over AI processing with enterprise-grade reliability and security.

📈 Next Steps: Enhanced monitoring, multi-proxy support, and advanced local AI model management coming in future releases.

v2.1.2

🚀 Agentic Markdown Todos v2.1.2

🏠 Major Feature: Local AI Proxy Implementation

Revolutionary AI Independence

This release introduces Local AI Proxy as a first-class AI provider, enabling complete data sovereignty and AI processing independence:

• 🔐 Complete Data Control: Process ALL AI requests locally without external service dependency
• 💰 Cost Elimination: Avoid external API costs with local AI infrastructure
• 🏢 Enterprise Compliance: Meet strict data governance requirements
• 🛠️ Multi-Platform Support: Compatible with LMStudio, Ollama, and custom OpenAI endpoints

Key Technical Features

• Provider-Based Architecture: Simple, reliable AI request routing
• Real-Time Connection Monitoring: Live status validation with 30-second health checks
• WebSocket Infrastructure: Efficient proxy communication system
• User-Specific Authentication: UUID + localToken security model
• Enhanced Setup UI: Tabbed configuration wizard with comprehensive guidance

Docker Integration

# LMStudio Configuration
docker run -d --name am_todos_proxy \
  -v ./proxy-config:/config \
  -e LOCAL_PROXY_MODE=true \
  -e MAIN_SERVER_TOKEN=<your-token> \
  -e MAIN_SERVER_URL=wss://your-app.run.app/proxy-ws \
  -e LOCAL_AI_ENDPOINT=http://host.docker.internal:1234 \
  -e LOCAL_AI_MODEL=mistralai/mistral-small-3.2 \
  ghcr.io/florinpeter/am-todos:latest

🐛 Critical Bug Fixes

Local Proxy Configuration Loop Fix

• Fixed: Infinite loop during "Proxy UUID" entry that prevented "Local Token" input
• Root Cause: Cascading useEffect dependencies creating dependency chain loops
• Solution: Enhanced validation guards and proper dependency management

Component Naming Consistency

• Updated: GitSettings.tsx → GeneralSettings.tsx for semantic accuracy
• Fixed: TypeScript import conflicts between component and type names
• Improved: All references updated across 13+ files for consistency

🏗️ Architecture Improvements

Simplified & Efficient Design

• 99% Reduction: Eliminated complex content analysis overhead
• Net Code Reduction: -89 lines across 17 files with enhanced functionality
• Modular Structure: New server architecture with dedicated route modules
• Enhanced Security: Multi-layer authentication and validation system

New Server Structure

server/
├── config/        # Application and proxy configuration
├── middleware/    # Authentication and security middleware  
├── routes/        # Modular API routes (ai, git, proxy, search)
├── services/      # Local proxy and proxy manager services
├── utils/         # Authentication utilities
└── websocket/     # WebSocket communication handlers

📊 Quality Assurance

Comprehensive Testing

• 1,017 Tests Passing: Complete test suite validation
• 79.4% Code Coverage: Extensive coverage across all components
• CI/CD Integration: Automated testing with GitHub Actions
• Multi-Platform Testing: Node.js 20.x and 22.x validation

Production Ready

• Zero Breaking Changes: Full backward compatibility maintained
• Enhanced Error Handling: Comprehensive error states with recovery guidance
• Performance Optimized: Faster routing decisions without content analysis
• Mobile Responsive: Complete mobile-first design with touch-friendly interface

🌟 Business Impact

For Organizations

• Data Sovereignty: Complete control over AI processing and data
• Compliance Ready: Meet strict enterprise security requirements
• Cost Control: Eliminate external AI API expenses
• Infrastructure Independence: Use existing local AI investments

For Users

• Seamless Experience: Local proxy works identically to cloud providers
• Real-Time Feedback: Live connection status and troubleshooting guidance
• Flexible Setup: Support for popular local AI platforms
• Enhanced Security: User-specific proxy routing with credential validation

📚 Documentation

New Documentation

• Local AI Proxy Guide: Complete implementation documentation
• Enhanced Development Setup: Improved restart scripts with test tokens
• Architecture Overview: Detailed technical implementation guide
• Security Model: Multi-layer authentication and authorization documentation

🔄 Migration Notes

Automatic Updates

• Seamless Upgrade: No breaking changes for existing users
• Settings Migration: Automatic localStorage updates
• Frontmatter Simplification: Reduced to tags-only structure
• Test Validation: All existing functionality preserved

New Users

1. Select AI Provider: Choose "Local Proxy" alongside Gemini and OpenRouter
2. Deploy Proxy: Use provided Docker commands for your AI platform
3. Configure Credentials: Copy UUID and localToken from proxy settings
4. Test Connection: Validate setup with real-time status monitoring
5. Start Using: All AI features work seamlessly with local processing

---

🏆 Complete Success: This release transforms AM-Todos into a truly data-sovereign AI-powered task management system, providing unprecedented control over AI processing while maintaining seamless user experience and enterprise-grade security.

📈 Next Steps: Enhanced monitoring, multi-proxy support, and advanced local AI model management coming in future releases.

v2.1.1

🚀 Agentic Markdown Todos v2.1.1

🏠 Major Feature: Local AI Proxy Implementation

Revolutionary AI Independence

This release introduces Local AI Proxy as a first-class AI provider, enabling complete data sovereignty and AI processing independence:

• 🔐 Complete Data Control: Process ALL AI requests locally without external service dependency
• 💰 Cost Elimination: Avoid external API costs with local AI infrastructure
• 🏢 Enterprise Compliance: Meet strict data governance requirements
• 🛠️ Multi-Platform Support: Compatible with LMStudio, Ollama, and custom OpenAI endpoints

Key Technical Features

• Provider-Based Architecture: Simple, reliable AI request routing
• Real-Time Connection Monitoring: Live status validation with 30-second health checks
• WebSocket Infrastructure: Efficient proxy communication system
• User-Specific Authentication: UUID + localToken security model
• Enhanced Setup UI: Tabbed configuration wizard with comprehensive guidance

Docker Integration

# LMStudio Configuration
docker run -d --name am_todos_proxy \
  -v ./proxy-config:/config \
  -e LOCAL_PROXY_MODE=true \
  -e MAIN_SERVER_TOKEN=<your-token> \
  -e MAIN_SERVER_URL=wss://your-app.run.app/proxy-ws \
  -e LOCAL_AI_ENDPOINT=http://host.docker.internal:1234 \
  -e LOCAL_AI_MODEL=mistralai/mistral-small-3.2 \
  ghcr.io/florinpeter/am-todos:latest

🐛 Critical Bug Fixes

Local Proxy Configuration Loop Fix

• Fixed: Infinite loop during "Proxy UUID" entry that prevented "Local Token" input
• Root Cause: Cascading useEffect dependencies creating dependency chain loops
• Solution: Enhanced validation guards and proper dependency management

Component Naming Consistency

• Updated: GitSettings.tsx → GeneralSettings.tsx for semantic accuracy
• Fixed: TypeScript import conflicts between component and type names
• Improved: All references updated across 13+ files for consistency

🏗️ Architecture Improvements

Simplified & Efficient Design

• 99% Reduction: Eliminated complex content analysis overhead
• Net Code Reduction: -89 lines across 17 files with enhanced functionality
• Modular Structure: New server architecture with dedicated route modules
• Enhanced Security: Multi-layer authentication and validation system

New Server Structure

server/
├── config/        # Application and proxy configuration
├── middleware/    # Authentication and security middleware  
├── routes/        # Modular API routes (ai, git, proxy, search)
├── services/      # Local proxy and proxy manager services
├── utils/         # Authentication utilities
└── websocket/     # WebSocket communication handlers

📊 Quality Assurance

Comprehensive Testing

• 1,017 Tests Passing: Complete test suite validation
• 79.4% Code Coverage: Extensive coverage across all components
• CI/CD Integration: Automated testing with GitHub Actions
• Multi-Platform Testing: Node.js 20.x and 22.x validation

Production Ready

• Zero Breaking Changes: Full backward compatibility maintained
• Enhanced Error Handling: Comprehensive error states with recovery guidance
• Performance Optimized: Faster routing decisions without content analysis
• Mobile Responsive: Complete mobile-first design with touch-friendly interface

🌟 Business Impact

For Organizations

• Data Sovereignty: Complete control over AI processing and data
• Compliance Ready: Meet strict enterprise security requirements
• Cost Control: Eliminate external AI API expenses
• Infrastructure Independence: Use existing local AI investments

For Users

• Seamless Experience: Local proxy works identically to cloud providers
• Real-Time Feedback: Live connection status and troubleshooting guidance
• Flexible Setup: Support for popular local AI platforms
• Enhanced Security: User-specific proxy routing with credential validation

📚 Documentation

New Documentation

• Local AI Proxy Guide: Complete implementation documentation
• Enhanced Development Setup: Improved restart scripts with test tokens
• Architecture Overview: Detailed technical implementation guide
• Security Model: Multi-layer authentication and authorization documentation

🔄 Migration Notes

Automatic Updates

• Seamless Upgrade: No breaking changes for existing users
• Settings Migration: Automatic localStorage updates
• Frontmatter Simplification: Reduced to tags-only structure
• Test Validation: All existing functionality preserved

New Users

1. Select AI Provider: Choose "Local Proxy" alongside Gemini and OpenRouter
2. Deploy Proxy: Use provided Docker commands for your AI platform
3. Configure Credentials: Copy UUID and localToken from proxy settings
4. Test Connection: Validate setup with real-time status monitoring
5. Start Using: All AI features work seamlessly with local processing

---

🏆 Complete Success: This release transforms AM-Todos into a truly data-sovereign AI-powered task management system, providing unprecedented control over AI processing while maintaining seamless user experience and enterprise-grade security.

📈 Next Steps: Enhanced monitoring, multi-proxy support, and advanced local AI model management coming in future releases.

v2.0.0

🚀 Agentic Markdown Todos v2.0.0

🎯 Major Features

✨ Enhanced Todo Creation with Templates

• Template System: 6 predefined templates for different task types (General, Project Planning, Bug Investigation, Feature Development, Research, Personal Goals)
• Structured AI Output: All AI actions now use consistent JSON format for reliable parsing
• Enhanced Input Interface: Title + optional description + template selection for better AI generation
• Backward Compatibility: Maintains compatibility with existing todo creation workflow

🏗️ Advanced UI & Performance Improvements

• Perfect Checkbox Alignment: Fixed checkbox positioning with CSS transforms for mobile and desktop
• Enhanced Markdown Rendering: GitHub-style display with improved visual consistency
• Filename-Based Metadata: 99%+ reduction in API requests through smart filename encoding
• Consolidated Test Suite: Reduced from 109 → 33 test files (69.7% reduction) while achieving 81% coverage

🛠️ Development Experience

• PostToolUse Hook System: Automated code quality enforcement with TypeScript and ESLint checks
• Comprehensive Testing: 1079 tests with anti-cluttering guidelines and coverage analysis
• Developer Tools: Enhanced scripts and hooks for quality assurance

📋 Complete Feature Set

Core Functionality

• ✅ AI-Powered Task Generation with multi-provider support (Google Gemini, OpenRouter)
• ✅ Interactive Markdown Editor with real-time Git sync
• ✅ Multi-Folder Support for project organization
• ✅ Intelligent Search with real-time results and keyboard shortcuts
• ✅ Configuration Sharing with QR codes for cross-device setup
• ✅ Mobile-Responsive Design with touch-friendly interface

Technical Excellence

• ✅ Type Safety: Comprehensive TypeScript implementation
• ✅ Performance Optimized: Smart caching and API request reduction
• ✅ Security Focused: Fine-grained GitHub/GitLab permissions
• ✅ Extensible Architecture: Plugin-ready template system

🔧 Technical Improvements

Enhanced AI Integration

• Multi-Provider Support: Choose between Google Gemini or OpenRouter (400+ models)
• Structured Output: Consistent JSON responses with fallback mechanisms
• Template-Guided Generation: Specialized AI prompts for different task categories
• Improved Reliability: Better error handling and response parsing

UI/UX Enhancements

• Perfect Alignment: Checkbox positioning fixed with translateY(4px) transform
• Template Selection: Intuitive dropdown with category descriptions
• Progressive Disclosure: Optional description field with toggle
• Visual Consistency: GitHub-style markdown rendering

Performance & Reliability

• Filename Metadata: Instant todo listing without individual file fetches
• Smart Caching: Optimized data loading and storage
• Error Recovery: Comprehensive error handling with user-friendly messages
• Test Coverage: 81.08% coverage with focused, high-quality tests

📈 Metrics

• Test Files: 109 → 33 (69.7% reduction, anti-cluttering success)
• Test Coverage: 81.08% across 1079 tests
• API Performance: 99%+ reduction in listing operations
• TypeScript: Zero compilation errors with strict checking
• Code Quality: All ESLint rules satisfied

🆕 What's New

1. Enhanced Todo Creation: Template-guided AI generation with structured output
2. Perfect Mobile Experience: Fixed checkbox alignment and responsive design
3. Advanced Development Tools: PostToolUse hooks and quality automation
4. Improved Performance: Filename-based metadata and smart caching
5. Better Testing: Consolidated test suite with comprehensive coverage

🔄 Migration Guide

No migration required! This release is fully backward compatible:

• Existing todos continue to work without changes
• Old creation workflow remains functional
• All existing features and functionality preserved
• New template features available immediately

📚 Documentation

• Enhanced Todo Creation Guide
• Filename-Based Metadata
• Testing Documentation
• Development Setup

---

🎉 Ready for Production: All features tested, documented, and production-ready!

💫 You Own Your Data: Tasks remain as .md files in your Git repository, editable with any tool.

v1.9.1

🎉 Release v1.9.1: Enhanced Git History with Priority Restoration Fixes

This release delivers comprehensive improvements to the Git History feature, combining critical bug fixes for priority restoration with rich metadata display capabilities.

🐛 Critical Bug Fixes

Priority Restoration Issues Resolved ✅

• Fixed priority loss during Git History restoration - Priority information now properly preserved when restoring from commit history
• Fixed priority loss during Checkpoint restoration - AI chat checkpoint restoration now maintains task priority
• Fixed priority loss during Draft restoration - Auto-saved drafts now preserve priority changes
• Enhanced frontmatter handling - All restoration mechanisms now properly handle YAML frontmatter

✨ New Features

Enhanced Git History with Rich Metadata Display

• Priority Badges (P1-P5) with intuitive color coding in commit list:
  • 🔴 P1: Critical (Red)
  • 🟠 P2: High (Orange)
  • 🟡 P3: Medium (Yellow)
  • 🔵 P4: Low (Blue)
  • ⚫ P5: Very Low (Gray)
• Archive Indicators (📁 ARCHIVED) for completed tasks
• Smart Date Display showing creation date when different from commit date
• Loading Indicators with smooth animations for better UX

Enhanced Preview Panel

• Metadata Section displaying priority, archive status, and task title
• Mobile-Responsive Design with consistent experience across devices
• Rich Context for making informed restoration decisions

⚡ Performance Improvements

Intelligent Caching System

• Frontmatter Caching with 5-minute expiry to avoid redundant parsing
• Upfront Loading of first 5 commits for immediate metadata display
• Memoized Helper Functions to prevent unnecessary re-renders
• Smart Cache Cleanup to prevent memory bloat (auto-cleanup after 100 entries)

🧪 Testing Enhancements

Comprehensive Test Coverage

• 19 total tests across priority restoration and metadata display
• Integration tests for complete priority workflows
• Performance tests for caching and optimization
• Mobile responsive tests for UI consistency
• Error handling tests for edge cases and malformed data

📱 User Experience Improvements

Before vs After

Before:

• Git History showed only basic commit information
• Priority information lost during restoration operations
• Users had to preview each commit to understand context
• No indication of task priority or archive status

After:

• At-a-glance context with priority and archive status in commit list
• Priority preservation across all restoration mechanisms
• Rich metadata in preview panel with task details
• Performance optimized with smart caching and loading strategies

🔧 Technical Implementation

Architecture Enhancements

• Extended GitCommit interface with optional frontmatter and loading state tracking
• Implemented getCachedFrontmatter() with content hash + commit SHA caching
• Added batch processing with loadFrontmatterForCommits() for efficient API usage
• Created memoized UI components for optimal rendering performance

New Components

• getPriorityBadge(): Color-coded priority indicators
• getArchiveBadge(): Archive status indicators
• formatCreatedDate(): Smart date formatting
• renderCommitMetadata(): Unified metadata rendering

🎯 Impact

This release transforms the Git History view from a basic commit browser into a powerful restoration tool that provides users with all the context they need to make informed decisions, while ensuring that priority information is never lost during any restoration operation.

📦 What's Included

• ✅ Priority restoration bug fixes
• ✅ Enhanced Git History with metadata display
• ✅ Performance optimizations with intelligent caching
• ✅ Mobile-responsive design improvements
• ✅ Comprehensive test coverage (19 tests)
• ✅ Full backward compatibility

🔗 Related

• Pull Request: #48 Enhanced Git History with metadata display and performance optimization
• Previous Release: v1.9.0 CodeMirror Editor

---

Full Changelog: v1.9.0...v1.9.1

🤖 Generated with Claude Code

v1.9.0

🎉 Agentic Markdown Todos v1.9.0

🚀 Major New Feature: Advanced Markdown Editor

This release introduces a completely redesigned markdown editing experience powered by CodeMirror 6, replacing the basic textarea with a professional code editor.

✨ New Editor Features

🔤 CodeMirror 6 Integration

• Syntax Highlighting: Full GitHub-flavored markdown syntax highlighting
• Dark Theme: Custom dark theme matching the app's design (gray-900 background)
• Auto-Completion: Intelligent markdown completion and auto-list continuation
• Enhanced Selection: Proper text selection with blue highlighting
• Responsive Design: Optimized for both desktop and mobile usage

😊 Slack-Style Emoji Support

• Shortcode Completion: Type :rocket: → 🚀, :+1: → 👍, :heart: → ❤️
• Popular Emojis: Instant access to commonly used emojis with autocomplete
• Smart Search: Emoji search with fallback handling for special characters
• Real-time Preview: See emoji suggestions as you type with character preview

🛠 Technical Improvements

• Better Performance: Optimized rendering and interaction handling
• Accessibility: Improved keyboard navigation and screen reader support
• Mobile Optimization: Touch-friendly interface with responsive design
• Error Handling: Robust error handling for edge cases and special characters

🔧 Under the Hood

Dependencies Added

• @codemirror/view: ^6.34.3 - Core editor view functionality
• @codemirror/state: ^6.4.1 - Editor state management
• @codemirror/lang-markdown: ^6.3.1 - Markdown language support
• @codemirror/autocomplete: ^6.18.3 - Autocompletion system
• @uiw/react-codemirror: ^4.23.8 - React integration wrapper
• node-emoji: ^2.1.3 - Emoji search and conversion

Testing Infrastructure

• 100% Test Coverage: Comprehensive test suite for all editor functionality
• JSDOM Compatibility: Custom DOM Range API mocking for CodeMirror in test environment
• CI/CD Integration: All tests passing on Node.js 20.x and 22.x
• 67+ New Tests: Covering editor functionality, emoji completion, and integration

Code Quality

• TypeScript Support: Full type safety for all new components
• Component Architecture: Clean separation between editor and markdown viewer
• Extension System: Modular emoji extension for future expandability
• Error Boundaries: Comprehensive error handling and fallback mechanisms

🎯 User Experience Enhancements

Improved Editing Workflow

• Visual Feedback: Real-time unsaved changes indicator
• Smart Auto-Lists: Automatic todo list continuation (- [ ] when pressing Enter)
• Keyboard Shortcuts: Standard editor shortcuts for copy/paste/select
• Undo/Redo: Built-in undo/redo functionality

Mobile & Accessibility

• Touch Support: Optimized touch interaction for mobile devices
• Screen Reader: Proper ARIA labels and semantic markup
• High Contrast: Improved color contrast ratios for better visibility
• Keyboard Navigation: Full keyboard accessibility support

🔄 Migration & Compatibility

• Seamless Upgrade: Existing markdown content works without changes
• Backward Compatible: All existing features and functionality preserved
• No Data Loss: Safe migration from textarea to CodeMirror editor
• Settings Preserved: All user preferences and configurations maintained

🐛 Bug Fixes & Improvements

• Text Selection: Fixed selection highlighting issues in edit mode
• Focus Management: Improved focus handling during mode switches
• Draft Persistence: Enhanced draft saving and restoration
• Error Recovery: Better error handling for edge cases

📈 Performance Improvements

• Faster Rendering: Optimized editor initialization and updates
• Memory Efficiency: Reduced memory footprint with efficient DOM handling
• Network Optimization: Bundled dependencies for faster loading
• Responsive UI: Smoother interactions and transitions

---

🛠 Development & Testing

This release includes a comprehensive testing infrastructure with 1,096+ passing tests across:

• Unit tests for all new components
• Integration tests for editor functionality
• End-to-end testing for user workflows
• Cross-platform compatibility testing (Node.js 20.x & 22.x)

🙏 Acknowledgments

Special thanks to the CodeMirror team for creating such an excellent editor framework, and to the open-source community for the emoji and testing libraries that make this release possible.

---

Full Changelog: v1.7.2...v1.9.0

v1.8.2

🚨 AM-Todos v1.8.2: Critical Hotfix - Cloud Run Deployment Fix

🐛 Critical Issue Fixed

This hotfix resolves a critical deployment failure introduced in v1.8.1 that prevented the application from starting in Cloud Run production environments.

⚡ What Was Broken

v1.8.1 Cloud Run deployments failed with:

The user-provided container failed to start and listen on the port defined provided by the PORT=8080 environment variable

🔍 Root Cause Analysis

Through Cloud Run logs investigation, we identified two separate issues:

1. Missing Module Error

Error [ERR_MODULE_NOT_FOUND]: Cannot find module '/app/server/utils/redosProtection.js' imported from /app/server/server.js

Cause: Dockerfile was only copying individual server files but missing the utils/ subdirectory

2. Server Startup Condition Failure

Cause: Path resolution inconsistency between import.meta.url and process.argv[1] in containerized environments

✅ Fixes Applied

📦 Fix 1: Dockerfile - Missing Utils Directory

# Added missing utils directory copy
COPY --chown=nextjs:nodejs server/utils/ ./server/utils/

🔧 Fix 2: Server Startup - Robust Path Resolution

// Before (broken in containers)
if (import.meta.url === `file://${process.argv[1]}`) {

// After (works everywhere)
import { pathToFileURL } from 'url';
if (import.meta.url === pathToFileURL(path.resolve(process.argv[1])).href) {

🧪 Verification

• ✅ Local development: Server starts correctly
• ✅ Container build: All required modules included
• ✅ Path resolution: Works in both development and production
• ✅ Cloud Run ready: Proper port binding and health checks

📊 Impact

• Severity: P1 Critical - Blocked all v1.8.1 production deployments
• Affected versions: v1.8.1 only
• Resolution: Complete - deployments now work correctly

🚀 Deployment Ready

This hotfix ensures:

• ✅ Container starts successfully - no missing modules
• ✅ Server listens on PORT - robust startup condition
• ✅ Health checks pass - proper application initialization
• ✅ Full functionality - all features work as expected

🔗 Technical Details

• PR: #45 - Cloud Run deployment failure fix
• Files changed: Dockerfile, server/server.js
• Commits: 2 targeted fixes for both root causes
• Backward compatibility: Maintained for development environments

📋 Upgrade Instructions

Deploy the new v1.8.2 image directly - no configuration changes needed:

export SOURCE_IMAGE="ghcr.io/florinpeter/am-todos:v1.8.2"
./hack/deploy-all.sh

---

Full Changelog: v1.8.1...v1.8.2

v1.8.1

🚀 AM-Todos v1.8.1: Structured AI Responses for Commit Messages

🎯 What's New

Robust AI Response Parsing for Commit Messages

• Multi-layer parsing strategy handles any AI response format
• Markdown code block extraction for responses like:
  Sure, here's a conventional commit message:

  fix: Update "Claude Code Setup" todo list item
  

• JSON extraction from ```json code blocks
• Smart cleanup removes AI prefixes automatically
• Enhanced logging shows exactly how responses are processed

Structured Response Benefits

• Consistent API patterns across all AI functions (generateInitialPlan, generateCommitMessage, processChatMessage)
• Richer user feedback during commit operations with meaningful descriptions
• Better debugging capabilities with comprehensive logging
• Future-proof adapts to AI model changes automatically

🔧 Technical Improvements

6-Step Parsing Strategy

1. Direct JSON parsing - {"message": "...", "description": "..."}
2. JSON from markdown - Extract from ```json blocks
3. Commit from code blocks - Extract commit messages from ``` blocks
4. Pattern matching - Find conventional commit patterns in text
5. General matching - Flexible commit-like pattern detection
6. Smart cleanup - Remove AI prefixes and clean response

Enhanced Logging & Monitoring

• Raw response logging (first 200 chars) for debugging
• Parsing method tracking shows which strategy succeeded
• Provider/model information for pattern analysis
• Backward compatibility maintained for existing functionality

Comprehensive Testing

• 8 new test cases covering all response formats
• Real-world scenarios tested including exact user cases
• Backward compatibility verified with plain text responses
• Integration tests for both Gemini and OpenRouter providers

📊 Statistics

• 9/9 generateCommitMessage tests passing
• 60/60 feature validation tests passing
• Zero breaking changes - full backward compatibility
• 290+ lines of new robust parsing logic

🎨 User Experience Improvements

Always Works

No more failed commit message generation - the system gracefully handles:

• Structured JSON responses
• Markdown with code blocks
• Plain text with AI prefixes
• Malformed or unexpected responses

Better Feedback

Users now see meaningful descriptions like:

• "Generated conventional commit message for new todo creation"
• "Extracted commit message from markdown response"
• "Used cleaned AI response as commit message"

🔗 Related Issues

• Fixes markdown response handling for AI commit messages
• Resolves parsing failures with AI prefix text
• Improves debugging capabilities for AI integrations
• Enhances consistency across AI service functions

---

Full Changelog: v1.8.0...v1.8.1

v1.8.0

🚀 Major Features

🤖 Structured AI Responses

• Meaningful Feedback: AI now provides detailed descriptions instead of generic "Task updated successfully" messages
• JSON Mode: Both Gemini and OpenRouter providers return structured responses with content + description
• Better UX: Users see exactly what AI did (e.g., "Added authentication step to task 3")

💬 Context-Aware AI Chat

• Natural Follow-ups: Use pronouns like "it", "that", "the last one" in conversations
• Iterative Refinement: Chain requests like "make it shorter" → "now make it formal"
• Conversation Memory: AI receives last 6 messages for optimal context understanding
• Stateless Backend: Chat history managed in frontend for privacy and performance

🧪 Testing & Quality Improvements

Performance

• 24% Faster Tests: New fast test configuration for development
• CI Stability: Fixed test timeouts and hanging issues
• 138 New Tests: Comprehensive coverage for new features

Coverage

• Improved Metrics: Better coverage for modified components
• Clean Reporting: Excluded TypeScript interface files from coverage
• Error Handling: Comprehensive tests for edge cases and error scenarios

🛠️ Technical Enhancements

Type Safety

• Structured Interfaces: TypeScript interfaces for AI responses
• Backward Compatibility: Graceful fallback to plain text responses
• Error Handling: Robust error handling with meaningful user feedback

Development Experience

• Fast Test Config: 24% performance improvement for development testing
• Clean Coverage: Focused coverage metrics excluding interface files
• Stable CI: Resolved hanging and timeout issues

🔧 Bug Fixes

• Fixed test timeouts in CI environment
• Resolved package dependency synchronization issues
• Improved error handling for network and API failures
• Enhanced markdown rendering edge cases

📋 Complete Changelog

• feat: implement structured AI responses for better user feedback
• feat: implement context-aware AI chat with conversation history
• test: improve coverage for files modified in this PR
• chore: exclude TypeScript interface files from coverage reporting
• fix: update AIChat test for new context-aware signature
• chore: clean up documentation and exclude fast config from coverage
• revert: restore original vitest.config.mjs to remove verbose output
• fix: update package-lock.json for happy-dom dependency
• feat: add fast test command with 24% performance improvement
• fix: update test expectations for structured AI responses

🎯 Migration Notes

• Automatic: All changes are backward compatible
• No Action Required: Existing configurations continue to work
• Enhanced UX: Users immediately benefit from improved AI feedback

---

This release significantly enhances the AI interaction experience while maintaining full backward compatibility and improving the development workflow with better testing and CI stability.