massgen 0.1.0a2__tar.gz → 0.1.8__tar.gz

{massgen-0.1.0a2 → massgen-0.1.8}/CHANGELOG.md

@@ -7,18 +7,511 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## Recent Releases
 
-**v0.1.0 (October 2025)** - PyPI Package & Documentation
-Official PyPI release with enhanced documentation and installation via `pip install massgen`.
+**v0.1.8 (November 5, 2025)** - Automation Mode & DSPy Integration
+Complete automation infrastructure for LLM agents with real-time status tracking, silent execution mode, and DSPy-powered question paraphrasing for enhanced multi-agent diversity.
 
-**v0.0.32 (October 2025)** - Docker Execution Mode
-Isolated command execution in Docker containers with resource limits and multi-agent support.
+**v0.1.7 (November 3, 2025)** - Agent Task Planning & Background Execution
+Agent task planning system with dependency tracking, background shell execution for long-running commands, and preemption-based coordination for improved multi-agent workflows.
 
-**v0.0.31 (October 2025)** - Code Execution, AG2 Integration, Multimodal
-MCP-based command execution across all backends, AG2 framework integration, and audio/video generation.
+**v0.1.6 (October 31, 2025)** - Framework Interoperability & Backend Refactoring
+External agent framework integration as tools (AG2, LangGraph, AgentScope, OpenAI Assistants, SmoLAgent), unified tool execution architecture, Gemini backend simplification, and comprehensive configuration validation.
 
 ---
 
-## [0.1.0] - 2025-10-15 (PyPI Release)
+## [0.1.8] - 2025-11-05
+
+### Added
+- **Automation Mode for LLM Agents**: Complete infrastructure for running MassGen via LLM agents and programmatic workflows
+  - New `--automation` CLI flag for silent execution with minimal output (~10 lines vs 250-3,000+)
+  - New `SilentDisplay` class in `massgen/frontend/displays/silent_display.py` for automation-friendly output
+  - Real-time `status.json` monitoring file updated every 2 seconds via enhanced `CoordinationTracker`
+  - Meaningful exit codes: 0 (success), 1 (config error), 2 (execution error), 3 (timeout), 4 (interrupted)
+  - Automatic workspace isolation for parallel execution with unique suffixes
+  - Meta-coordination capabilities: MassGen running MassGen configurations
+  - Automatic log directory creation and management for automation sessions
+
+- **DSPy Question Paraphrasing Integration**: Intelligent question diversity for multi-agent coordination
+  - New `massgen/dspy_paraphraser.py` module with semantic-preserving paraphrasing (557 lines)
+  - Three paraphrasing strategies: "diverse", "balanced" (default), "conservative"
+  - Configurable number of variants per orchestrator session
+  - Automatic semantic validation using `SemanticValidationSignature` to ensure meaning preservation
+  - Thread-safe caching system with SHA-256 hashing for performance
+  - Support for all backends (Gemini, OpenAI, Claude, etc.) as paraphrasing engines
+
+- **Case Study Summary**: Comprehensive documentation of MassGen capabilities
+  - New `docs/CASE_STUDIES_SUMMARY.md` providing centralized overview of 33 case studies (368 lines)
+  - Organized by category: Release Features, Research, Travel, Creative, In Development, Planned
+  - Covers versions v0.0.3 to v0.1.5 with status tracking and links to videos
+  - Statistics: 19 completed, 8 with video demonstrations, 6 categories
+
+### Changed
+- **Orchestrator Enhancement**: Integration of DSPy paraphrasing and automation tracking
+  - Question variant distribution to different agents based on configured strategy
+  - Improved coordination event logging with structured status exports
+
+- **CLI Enhancement**: Extended command-line interface for automation workflows
+  - Enhanced display selection logic automatically choosing SilentDisplay in automation mode
+  - Improved output formatting optimized for LLM agent parsing and monitoring
+
+### Documentations, Configurations and Resources
+
+- **Case Study**: Meta-level self-analysis demonstrating automation mode
+  - New `docs/source/examples/case_studies/meta-self-analysis-automation-mode.md`: Comprehensive case study showing MassGen analyzing its own v0.1.8 codebase using automation mode
+
+- **Automation Documentation**: Comprehensive guides for LLM agent integration
+  - New `AI_USAGE.md`: Complete guide for LLM agents running MassGen (319 lines)
+  - New `docs/source/user_guide/automation.rst`: Full automation guide with BackgroundShellManager patterns (890 lines)
+  - New `docs/source/reference/status_file.rst`: Complete `status.json` schema reference with field-by-field documentation (565 lines)
+  - Updated `README.md` and `README_PYPI.md` with automation mode sections (135 lines each)
+
+- **DSPy Documentation**: Complete implementation and usage guide
+  - New `massgen/backend/docs/DSPY_IMPLEMENTATION_GUIDE.md`: Comprehensive DSPy integration guide (653 lines)
+  - Covers quick start, configuration, strategies, troubleshooting, and semantic validation
+  - Includes paraphrasing examples and best practices
+
+- **Meta-Coordination Configurations**: MassGen running MassGen examples
+  - `massgen/configs/meta/massgen_runs_massgen.yaml`: Single agent autonomously running MassGen experiments
+  - `massgen/configs/meta/massgen_suggests_to_improve_massgen.yaml`: Self-improvement configuration
+  - Demonstrates automation mode usage for meta-coordination workflows
+
+- **DSPy Configuration Example**: New YAML configuration for DSPy-enabled coordination
+  - `massgen/configs/basic/multi/three_agents_dspy_enabled.yaml`: Three-agent setup with DSPy paraphrasing
+
+- **Case Study Summary Documentation**: Centralized case study reference
+  - New `docs/CASE_STUDIES_SUMMARY.md`: Comprehensive overview of all MassGen case studies with categorization and status tracking
+
+### Technical Details
+- **Major Focus**: Automation infrastructure for LLM agents, DSPy-powered question paraphrasing, meta-coordination capabilities, comprehensive case study documentation
+- **Contributors**: @ncrispino @praneeth999 @franklinnwren @qidanrui @sonichi @Henry-811 and the MassGen team
+
+## [0.1.7] - 2025-11-03
+
+### Added
+- **Agent Task Planning System**: MCP-based task management with dependency tracking
+  - New `massgen/mcp_tools/planning/` module with dedicated planning server (`_planning_mcp_server.py`)
+  - Task dataclasses with dependency validation and status management (`planning_dataclasses.py`)
+  - Support for task states (pending/in_progress/completed/blocked) with automatic transitions based on dependencies
+  - Orchestrator integration for plan-aware coordination
+  - Test suite in `test_planning_integration.py` and `test_planning_tools.py`
+
+- **Background Shell Execution**: Long-running command support with persistent sessions
+  - New `BackgroundShell` class in `massgen/filesystem_manager/background_shell.py`
+  - Shell lifecycle management with output streaming and real-time monitoring
+  - Automatic timeout handling for long-running processes
+  - Enhanced code execution server with background execution capabilities
+  - Test coverage in `test_background_shell.py`
+
+- **Preemption Coordination**: Multi-agent coordination with interruption support
+  - Agents can preempt ongoing coordination to submit better answers without full restart
+  - Enhanced coordination tracker with preemption event logging
+  - Improved orchestrator logic to preserve partial progress during preemption
+
+### Fixed
+- **System Message Handling**: Resolved system message extraction in Claude Code backend for background shell execution
+- **Case Study Documentation**: Fixed broken links and outdated examples in older case studies
+
+### Documentations, Configurations and Resources
+
+- **Documentation Updates**: New user guides and design documentation
+  - New `docs/source/user_guide/agent_task_planning.rst`: Task planning guide with usage patterns and API reference
+  - Updated `docs/source/user_guide/code_execution.rst`: Added 122 lines for background shell usage
+  - New `docs/dev_notes/agent_planning_coordination_design.md`: Comprehensive design document for agent planning and coordination system
+  - New `docs/dev_notes/preempt_not_restart_design.md`: 456-line design document with preemption algorithms
+  - Updated `docs/source/development/architecture.rst`: Added 61 lines for preemption coordination architecture
+
+- **Configuration Examples**: New YAML configurations demonstrating v0.1.7 features
+  - `example_task_todo.yaml`: Task planning configuration
+  - `background_shell_demo.yaml`: Background shell execution demonstration
+
+### Technical Details
+- **Major Focus**: Agent task planning with dependencies, background command execution, preemption-based coordination
+- **Contributors**: @ncrispino @Henry-811 and the MassGen team
+
+## [0.1.6] - 2025-10-31
+
+### Added
+- **Framework Interoperability**: External agent framework integration as MassGen custom tools
+  - New `massgen/tool/_extraframework_agents/` module with 5 framework integrations
+  - **AG2 Lesson Planner Tool**: Nested chat functionality wrapped as custom tool for multi-agent lesson planning (supports streaming)
+  - **LangGraph Lesson Planner Tool**: LangGraph graph-based workflows integrated as tool
+  - **AgentScope Lesson Planner Tool**: AgentScope agent system wrapped for lesson creation
+  - **OpenAI Assistants Lesson Planner Tool**: OpenAI Assistants API integrated as tool
+  - **SmoLAgent Lesson Planner Tool**: HuggingFace SmoLAgent integration for lesson planning
+  - Enables MassGen agents to delegate tasks to specialized external frameworks
+  - Each framework runs autonomously and returns results to MassGen orchestrator
+  - Note: Only AG2 currently supports streaming; other frameworks return complete results
+
+- **Configuration Validator**: Comprehensive YAML configuration validation system
+  - New `ConfigValidator` class in `massgen/config_validator.py` for pre-flight validation
+  - Memory configuration validation with detailed error messages
+  - Pre-commit hook integration for automatic config validation
+  - Comprehensive test suite in `massgen/tests/test_config_validator.py`
+  - Validates agent configurations, backend parameters, tool settings, and memory options
+  - Provides actionable error messages with suggestions for common mistakes
+
+### Changed
+- **Backend Architecture Refactoring**: Unified tool execution with ToolExecutionConfig
+  - New `ToolExecutionConfig` dataclass in `base_with_custom_tool_and_mcp.py` for standardized tool handling
+  - Refactored `ResponseBackend` with unified tool execution flow
+  - Refactored `ChatCompletionsBackend` with unified tool execution flow
+  - Refactored `ClaudeBackend` with unified tool execution methods
+  - Eliminates duplicate code paths between custom tools and MCP tools
+  - Consistent error handling and status reporting across all tool types
+  - Improved maintainability and extensibility for future tool systems
+
+- **Gemini Backend Simplification**: Major architectural cleanup and consolidation
+  - Removed `gemini_mcp_manager.py` module
+  - Removed `gemini_trackers.py` module
+  - Refactored `gemini.py` to use manual tool execution via base class
+  - Streamlined tool handling and cleanup logic
+  - Removed continuation logic and duplicate code
+  - Updated `_gemini_formatter.py` for simplified tool conversion
+  - Net reduction of 1,598 lines through consolidation
+  - Improved maintainability and performance
+
+- **Custom Tool System Enhancement**: Improved tool management and execution
+  - Enhanced `ToolManager` with category management capabilities
+  - Improved tool registration and validation system
+  - Enhanced tool result handling and error reporting
+  - Better support for async tool execution
+  - Improved tool schema generation for LLM consumption
+
+### Documentations, Configurations and Resources
+
+- **Framework Interoperability Examples**: 8 new configuration files demonstrating external framework integration
+  - **AG2 Examples**: `ag2_lesson_planner_example.yaml`, `ag2_and_langgraph_lesson_planner.yaml`, `ag2_and_openai_assistant_lesson_planner.yaml`
+  - **LangGraph Examples**: `langgraph_lesson_planner_example.yaml`
+  - **AgentScope Examples**: `agentscope_lesson_planner_example.yaml`
+  - **OpenAI Assistants Examples**: `openai_assistant_lesson_planner_example.yaml`
+  - **SmoLAgent Examples**: `smolagent_lesson_planner_example.yaml`
+  - **Multi-Framework Examples**: `two_models_with_tools_example.yaml`
+
+### Technical Details
+- **Major Focus**: Framework interoperability for external agent integration, unified tool execution architecture, Gemini backend simplification, and configuration validation system
+- **Contributors**: @Eric-Shang @praneeth999 @ncrispino @qidanrui @sonichi @Henry-811 and the MassGen team
+
+## [0.1.5] - 2025-10-29
+
+### Added
+- **Memory System**: Complete long-term memory implementation with semantic retrieval
+  - New `massgen/memory/` module with comprehensive memory management
+  - **PersistentMemory** via mem0 integration for semantic fact storage and retrieval
+  - **ConversationMemory** for short-term verbatim message tracking
+  - **Automatic Context Compression** when approaching token limits
+  - **Memory Sharing for Multi-Turn Conversations** with turn-aware filtering to prevent temporal leakage
+  - **Session Management** for memory isolation and continuation across runs
+  - **Qdrant Vector Database Integration** for efficient semantic search (server and local modes)
+  - **Context Monitoring** with real-time token usage tracking
+  - Fact extraction prompts with customizable LLM and embedding providers
+  - Supports OpenAI, Anthropic, Groq, and other mem0-compatible providers
+
+- **Memory Configuration Support**: New YAML configuration options
+  - Memory enable/disable toggle at global and per-agent levels
+  - Configurable compression thresholds (trigger_threshold, target_ratio)
+  - Retrieval settings (limit, exclude_recent for smart retrieval)
+  - Session naming for continuation and cross-session memory
+  - LLM and embedding provider configuration for mem0
+  - Qdrant connection settings (server/local mode, host, port, path)
+
+### Changed
+- **Chat Agent Enhancement**: Memory integration for agent workflows
+  - Memory recording after agent responses (conversation and persistent)
+  - Memory retrieval on restart/reset for context restoration
+  - Integration with compression and context monitoring modules
+
+- **Orchestrator Enhancement**: Memory coordination for multi-agent workflows
+  - Memory initialization and management across agent lifecycles
+  - Memory cleanup on orchestrator shutdown
+
+### Documentations, Configurations and Resources
+
+- **Memory Documentation**: Comprehensive memory system user guide
+  - New `docs/source/user_guide/memory.rst`
+  - Complete usage guide with quick start, configuration reference, and examples
+  - Design decisions documentation explaining architecture choices
+  - Troubleshooting guide for common memory issues
+  - Monitoring and debugging instructions with log examples
+  - API reference for PersistentMemory, ConversationMemory, and ContextMonitor
+
+- **Configuration Examples**: 5 new memory-focused YAML configurations
+  - `gpt5mini_gemini_context_window_management.yaml`: Multi-agent with context compression
+  - `gpt5mini_gemini_research_to_implementation.yaml`: Research to implementation workflow
+  - `gpt5mini_high_reasoning_gemini.yaml`: High reasoning agents with memory
+  - `gpt5mini_gemini_baseline_research_to_implementation.yaml`: Baseline research workflow
+  - `single_agent_compression_test.yaml`: Testing compression behavior
+
+- **Infrastructure and Testing**:
+  - Memory test suite with 4 test files in `massgen/tests/memory/`
+  - Additional memory tests: `test_agent_memory.py`, `test_conversation_memory.py`, `test_orchestrator_memory.py`, `test_persistent_memory.py`
+
+### Technical Details
+- **Major Focus**: Long-term memory system with semantic retrieval and memory sharing for multi-turn conversations
+- **Contributors**: @ncrispino @qidanrui @kitrakrev @sonichi @Henry-811 and the MassGen team
+
+## [0.1.4] - 2025-10-27
+
+### Added
+- **Multimodal Generation Tools**: Comprehensive generation capabilities via OpenAI APIs
+  - New `text_to_image_generation` tool for generating images from text prompts using DALL-E models
+  - New `text_to_video_generation` tool for generating videos from text prompts
+  - New `text_to_speech_continue_generation` tool for text-to-speech with continuation support
+  - New `text_to_speech_transcription_generation` tool for audio transcription and generation
+  - New `text_to_file_generation` tool for generating documents (PDF, DOCX, XLSX, PPTX)
+  - New `image_to_image_generation` tool for image-to-image transformations
+  - Implemented in `massgen/tool/_multimodal_tools/` with 6 new modules
+
+- **Binary File Protection System**: Enhanced security for file operations
+  - New binary file blocking in `PathPermissionManager` preventing text tools from reading binary files
+  - Added `BINARY_FILE_EXTENSIONS` set covering images, videos, audio, archives, executables, and Office documents
+  - New `_validate_binary_file_access()` method with intelligent tool suggestions
+  - Prevents context pollution by blocking Read, read_text_file, and read_file tools from binary files
+  - Comprehensive test suite in `test_binary_file_blocking.py`
+
+- **Crawl4AI Web Scraping Integration**: Advanced web content extraction tool
+  - New `crawl4ai_tool` for intelligent web scraping with LLM-powered extraction
+  - Implemented in `massgen/tool/_web_tools/crawl4ai_tool.py`
+
+### Changed
+- **Multimodal File Size Limits**: Enhanced validation and automatic handling
+  - Automatic image resizing for files exceeding size limits
+  - Comprehensive size limit test suite in `test_multimodal_size_limits.py`
+  - Enhanced validation in understand_audio and understand_video tools
+
+### Documentations, Configurations and Resources
+
+- **PyPI Package Documentation**: Standalone README for PyPI distribution
+  - New `README_PYPI.md` with comprehensive package documentation
+  - Improved package metadata and installation instructions
+
+- **Release Management Documentation**: Comprehensive release workflow guide
+  - New `docs/dev_notes/release_checklist.md` with step-by-step release procedures
+  - Detailed checklist for testing, documentation, and deployment
+
+- **Binary File Protection Documentation**: Enhanced protected paths user guide
+  - Updated `docs/source/user_guide/protected_paths.rst` with binary file protection section
+  - Documents 40+ protected binary file types and specialized tool suggestions
+
+- **Configuration Examples**: 9 new YAML configuration files
+  - **Generation Tools**: 8 multimodal generation configurations
+    - `text_to_image_generation_single.yaml` and `text_to_image_generation_multi.yaml`
+    - `text_to_video_generation_single.yaml` and `text_to_video_generation_multi.yaml`
+    - `text_to_speech_generation_single.yaml` and `text_to_speech_generation_multi.yaml`
+    - `text_to_file_generation_single.yaml` and `text_to_file_generation_multi.yaml`
+  - **Web Scraping**: `crawl4ai_example.yaml` for Crawl4AI integration
+
+### Technical Details
+- **Major Focus**: Multimodal generation tools, binary file protection system, web scraping integration
+- **Contributors**: @qidanrui @ncrispino @sonichi @Henry-811 and the MassGen team
+
+## [0.1.3] - 2025-10-24
+
+### Added
+- **Post-Evaluation Workflow Tools**: Submit and restart capabilities for winning agents
+  - New `PostEvaluationToolkit` class in `massgen/tool/workflow_toolkits/post_evaluation.py`
+  - `submit` tool for confirming final answers
+  - `restart_orchestration` tool for restarting with improvements and feedback
+  - Post-evaluation phase where winning agent evaluates its own answer
+  - Support for all API formats (Claude, Response API, Chat Completions)
+  - Configuration parameter `enable_post_evaluation_tools` for opt-in/out
+
+- **Custom Multimodal Understanding Tools**: Active tools for analyzing workspace files using OpenAI's GPT-4.1 API
+  - New `understand_image` tool for analyzing images (PNG, JPEG, JPG) with detailed metadata extraction
+  - New `understand_audio` tool for transcribing and analyzing audio files (WAV, MP3, FLAC, OGG)
+  - New `understand_video` tool for extracting frames and analyzing video content (MP4, AVI, MOV, WEBM)
+  - New `understand_file` tool for processing documents (PDF, DOCX, XLSX, PPTX) with text and metadata extraction
+  - Works with any backend (uses OpenAI for analysis)
+  - Returns structured JSON with comprehensive metadata
+
+- **Docker Sudo Mode**: Enhanced Docker execution with privileged command support
+  - New `use_sudo` parameter for Docker execution
+  - Sudo mode for commands requiring elevated privileges
+  - Enhanced security instructions and documentation
+  - Test coverage in `test_code_execution.py`
+
+### Changed
+- **Interactive Config Builder Enhancement**: Improved workflow and provider handling
+  - Better flow from automatic setup to config builder
+  - Auto-detection of environment variables
+  - Improved provider-specific configuration handling
+  - Integrated multimodal tools selection in config wizard
+
+### Fixed
+- **System Message Warning**: Resolved deprecated system message configuration warning
+  - Fixed system message handling in `agent_config.py`
+  - Updated chat agent to properly handle system messages
+  - Removed deprecated warning messages
+
+- **Config Builder Issues**: Multiple configuration builder improvements
+  - Fixed config display errors
+  - Improved config saving across different provider types
+  - Better error handling for missing configurations
+
+### Documentations, Configurations and Resources
+
+- **Multimodal Tools Documentation**: Comprehensive documentation for new multimodal tools
+  - `docs/source/user_guide/multimodal.rst`: Updated with custom tools section
+  - `massgen/tool/docs/multimodal_tools.md`: Complete 779-line technical documentation
+
+- **Docker Sudo Mode Documentation**: Enhanced Docker execution documentation
+  - `docs/source/user_guide/code_execution.rst`: Added 98 lines documenting sudo mode
+  - `massgen/docker/README.md`: Updated with sudo mode instructions
+
+- **Configuration Examples**: New example configurations
+  - `configs/tools/multimodal_tools/understand_image.yaml`: Image analysis configuration
+  - `configs/tools/multimodal_tools/understand_audio.yaml`: Audio transcription configuration
+  - `configs/tools/multimodal_tools/understand_video.yaml`: Video analysis configuration
+  - `configs/tools/multimodal_tools/understand_file.yaml`: Document processing configuration
+
+- **Example Resources**: New test resources for v0.1.3 features
+  - `massgen/configs/resources/v0.1.3-example/multimodality.jpg`: Image example
+  - `massgen/configs/resources/v0.1.3-example/Sherlock_Holmes.mp3`: Audio example
+  - `massgen/configs/resources/v0.1.3-example/oppenheimer_trailer_1920.mp4`: Video example
+  - `massgen/configs/resources/v0.1.3-example/TUMIX.pdf`: PDF document example
+
+- **Case Studies**: New case study demonstrating v0.1.3 features
+  - `docs/source/examples/case_studies/multimodal-case-study-video-analysis.md`: Meta-level demonstration of multimodal video understanding with agents analyzing their own case study videos
+
+### Technical Details
+- **Major Focus**: Post-evaluation workflow tools, custom multimodal understanding tools, Docker sudo mode
+- **Contributors**: @ncrispino @qidanrui @sonichi @Henry-811 and the MassGen team
+
+## [0.1.2] - 2025-10-22
+
+### Added
+- **Claude 4.5 Haiku Support**: Added latest Claude Haiku model
+  - New model: `claude-haiku-4-5-20251001`
+  - Updated model registry in `backend/capabilities.py`
+
+### Changed
+- **Planning Mode Enhancement**: Intelligent automatic MCP tool blocking based on operation safety
+  - New `_analyze_question_irreversibility()` method in orchestrator analyzes questions to determine if MCP operations are reversible
+  - New `set_planning_mode_blocked_tools()`, `get_planning_mode_blocked_tools()`, and `is_mcp_tool_blocked()` methods in backend for selective tool control
+  - Dynamically enables/disables planning mode - read-only operations allowed during coordination, write operations blocked
+  - Planning mode supports different workspaces without conflicts
+  - Zero configuration required - works transparently
+
+
+- **Claude Model Priority**: Reorganized model list in capabilities registry
+  - Changed default model from `claude-sonnet-4-20250514` to `claude-sonnet-4-5-20250929`
+  - Moved `claude-opus-4-1-20250805` higher in priority order
+  - Updated in both Claude and Claude Code backends
+
+### Fixed
+- **Grok Web Search**: Resolved web search functionality in Grok backend
+  - Fixed `extra_body` parameter handling for Grok's Live Search API
+  - New `_add_grok_search_params()` method for proper search parameter injection
+  - Enhanced `_stream_with_custom_and_mcp_tools()` to support Grok-specific parameters
+  - Improved error handling for conflicting search configurations
+  - Better integration with Chat Completions API params handler
+
+### Documentations, Configurations and Resources
+
+- **Intelligent Planning Mode Case Study**: Complete feature documentation
+  - `docs/source/examples/case_studies/INTELLIGENT_PLANNING_MODE.md`: Comprehensive guide for automatic planning mode
+  - Demonstrates automatic irreversibility detection
+  - Shows read/write operation classification
+  - Includes examples for Discord, filesystem, and Twitter operations
+
+- **Configuration Updates**: Enhanced YAML examples
+  - Updated 5 planning mode configurations in `configs/tools/planning/` with selective blocking examples
+  - Updated `three_agents_default.yaml` with Grok-4-fast model
+  - Test coverage in `test_intelligent_planning_mode.py`
+
+### Technical Details
+- **Major Focus**: Intelligent planning mode with selective tool blocking, model support enhancements
+- **Contributors**: @franklinnwren @ncrispino @qidanrui @sonichi @Henry-811 and the MassGen team
+
+## [0.1.1] - 2025-10-20
+
+### Added
+- **Custom Tools System**: Complete framework for registering and executing user-defined Python functions as tools
+  - New `ToolManager` class in `massgen/tool/_manager.py` for centralized tool registration and lifecycle management
+  - Support for custom tools alongside MCP servers across all backends (Claude, Gemini, OpenAI Response API, Chat Completions, Claude Code)
+  - Three tool categories: builtin, mcp, and custom tools
+  - Automatic tool discovery with name prefixing and conflict resolution
+  - Tool validation with parameter schema enforcement
+  - Comprehensive test coverage in `test_custom_tools.py`
+
+- **Voting Sensitivity & Answer Novelty Controls**: Three-tier system for multi-agent coordination
+  - New `voting_sensitivity` parameter with three levels: "lenient", "balanced", "strict"
+  - "Lenient": Accepts any reasonable answer
+  - "Balanced": Default middle ground
+  - "Strict": High-quality requirement
+  - Answer novelty detection with `_check_answer_novelty()` method in `orchestrator.py` preventing duplicate answers
+  - Configurable `max_new_answers_per_agent` limiting submissions per agent
+  - Token-based similarity thresholds (50-70% overlap) for duplicate detection
+
+- **Interactive Configuration Builder**: Wizard for creating YAML configurations
+  - New `config_builder.py` module with step-by-step prompts
+  - Guided workflow for backend selection, model configuration, and API key setup
+  - Model-specific parameter handling (temperature, reasoning, verbosity)
+  - Tool enablement options (MCP servers, custom tools, builtin tools)
+  - Configuration validation and preview before saving
+  - Integration with `massgen --config-builder` command
+
+- **Backend Capabilities Registry**: Centralized feature support tracking
+  - New `capabilities.py` module in `massgen/backend/` documenting backend capabilities
+  - Feature matrix showing MCP, custom tools, multimodal, and code execution support
+  - Runtime capability queries for backend selection
+
+### Changed
+- **Gemini Backend Architecture**: Major refactoring for improved maintainability
+  - Extracted MCP management into `gemini_mcp_manager.py`
+  - Extracted tracking logic into `gemini_trackers.py`
+  - Extracted utilities into `gemini_utils.py`
+  - New API params handler `_gemini_api_params_handler.py`
+  - Improved session management and tool execution flow
+
+- **Python Version Requirements**: Updated minimum supported version
+  - Changed from Python 3.10+ to Python 3.11+ in `pyproject.toml`
+  - Ensures compatibility with modern type hints and async features
+
+- **API Key Setup Command**: Simplified command name
+  - Renamed `massgen --setup-keys` to `massgen --setup` for brevity
+  - Maintained all functionality for interactive API key configuration
+
+- **Configuration Examples**: Updated example commands
+  - Changed from `python -m massgen.cli` to simplified `massgen` command
+  - Updated 40+ configuration files for consistency
+
+### Fixed
+- **CLI Configuration Selection**: Resolved error with large config lists
+  - Fixed crash when using `massgen --select` with many available configurations
+  - Improved pagination and display of configuration options
+  - Enhanced error handling for configuration discovery
+
+- **CLI Help System**: Improved documentation display
+  - Fixed help text formatting in `massgen --help`
+  - Better organization of command options and examples
+
+### Documentations, Configurations and Resources
+
+- **Case Study: Universal Code Execution via MCP**: Comprehensive v0.0.31 feature documentation
+  - `docs/source/examples/case_studies/universal-code-execution-mcp.md`
+  - Demonstrates pytest test creation and execution across backends
+  - Shows command validation, security layers, and result interpretation
+
+- **Documentation Updates**: Enhanced existing documentation
+  - Added custom tools user guide and integration examples
+  - Reorganized case studies for improved navigation
+  - Updated configuration schema with new voting and tools parameters
+
+- **Custom Tools Examples**: 40+ example configurations
+  - Basic single-tool setups for each backend
+  - Multi-agent configurations with custom tools
+  - Integration examples combining MCP and custom tools
+  - Located in `configs/tools/custom_tools/`
+
+- **Voting Sensitivity Examples**: Configuration examples for voting controls
+  - `configs/voting/gemini_gpt_voting_sensitivity.yaml`
+  - Demonstrates lenient, balanced, and strict voting modes
+  - Shows answer novelty threshold configuration
+
+### Technical Details
+- **Major Focus**: Custom tools system, voting sensitivity controls, interactive config builder, and comprehensive documentation
+- **Contributors**: @qidanrui @ncrispino @praneeth999 @sonichi @Eric-Shang @Henry-811 and the MassGen team
+
+## [0.1.0] - 2025-10-17 (PyPI Release)
 
 ### Added
 - **PyPI Package Release**: Official MassGen package available on PyPI for easy installation via pip