This major refactoring addresses critical bugs in conversation history management
and significantly improves token efficiency through intelligent file embedding:
**Key Improvements:**
• Fixed conversation history duplication bug by centralizing reconstruction in server.py
• Added intelligent file filtering to prevent re-embedding files already in conversation history
• Centralized file processing logic in BaseTool._prepare_file_content_for_prompt()
• Enhanced log monitoring with better categorization and file embedding visibility
• Updated comprehensive test suite to verify new architecture and edge cases
**Architecture Changes:**
• Removed duplicate conversation history reconstruction from tools/base.py
• Conversation history now handled exclusively by server.py:reconstruct_thread_context
• All tools now use centralized file processing with automatic deduplication
• Improved token efficiency by embedding unique files only once per conversation
**Performance Benefits:**
• Reduced token usage through smart file filtering
• Eliminated redundant file embeddings in continued conversations
• Better observability with detailed debug logging for file operations
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
This commit addresses several critical issues and improvements:
🔧 Critical Fixes:
- Fixed conversation history not being included when using continuation_id in AI-to-AI conversations
- Fixed test mock targeting issues preventing proper conversation memory validation
- Fixed Docker debug logging functionality with Gemini tools
🐛 Bug Fixes:
- Docker compose configuration for proper container command execution
- Test mock import targeting from utils.conversation_memory.* to tools.base.*
- Version bump to 3.1.0 reflecting significant improvements
🚀 Improvements:
- Enhanced Docker environment configuration with comprehensive logging setup
- Added cross-tool continuation documentation and examples in README
- Improved error handling and validation across all tools
- Better logging configuration with LOG_LEVEL environment variable support
- Enhanced conversation memory system documentation
🧪 Testing:
- Added comprehensive conversation history bug fix tests
- Added cross-tool continuation functionality tests
- All 132 tests now pass with proper conversation history validation
- Improved test coverage for AI-to-AI conversation threading
✨ Code Quality:
- Applied black, isort, and ruff formatting across entire codebase
- Enhanced inline documentation for conversation memory system
- Cleaned up temporary files and improved repository hygiene
- Better test descriptions and coverage for critical functionality
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
• Moved AI-to-AI Conversation Threading section after "Start Using It\!"
• Now appears before Available Tools for logical progression
• Users learn basics first, then discover advanced collaboration features
• Improves documentation flow and user experience
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Major release introducing true AI collaboration capabilities:
• AI-to-AI conversation threading with Redis persistence
• Asynchronous workflow support - Claude can work between exchanges
• Incremental updates automatically bypass MCP's 25K token limit
• Cross-questioning and coordinated problem-solving between AIs
• Enhanced documentation with SwiftUI vs UIKit debate example
• Comprehensive test coverage and linting compliance
• Docker-first architecture with simplified setup
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Remove Windows batch scripts and native setup instructions since Claude Code CLI
requires WSL on Windows. Consolidate Docker as primary recommendation across all platforms.
Changes:
- Remove setup.bat, run_gemini.bat, setup-docker-env.bat
- Remove examples/claude_config_windows.json
- Update README to clarify WSL requirement for Windows users
- Promote Docker as recommended setup for all platforms
- Update troubleshooting section for WSL-only support
- Apply code formatting fixes from ruff/black
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Update DEBUG_ISSUE_PROMPT to request files when diagnostics are irrelevant/incomplete
- Clarify question field to specify what information is needed from Claude/user
- Update chat tool format_response to ask Claude to continue with user's request
- Update debug tool format_response to encourage implementing fixes when root cause identified
- Fix line length formatting to comply with 120 char limit
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add file request capability using JSON format like THINKDEEP_PROMPT
- Include technology stack awareness and grounding instructions
- Ensure collaboration stays within project scope and constraints
- Fix trailing whitespace issues identified by ruff linter
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add technology-specific considerations to all tool prompts (THINKDEEP, CODEREVIEW, DEBUG, ANALYZE, PRECOMMIT)
- Replace specific technology lists with generic key areas to prevent misdirection
- Add comprehensive grounding statements to prevent overstepping scope
- Enhance DEBUG_ISSUE_PROMPT with regression prevention guidance
- Add scope discipline and focus requirements to all prompts
- Fix trailing whitespace issues
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Add comprehensive technology-specific review categories and instructions to analyze codebase technology stack before reviewing. This enhancement will provide more targeted and relevant code reviews by adapting to different frameworks and patterns.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Fixed code formatting in server.py using black
- Only formatting changes, no functional modifications
- Ensures consistent code style across the project
All tests pass (96/96) and linting is clean:
- ✅ pytest: 96 tests passed
- ✅ ruff: all checks passed
- ✅ black: formatting applied
- ✅ isort: import sorting correct
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Replaced plain dict capabilities with proper ServerCapabilities object
- Added proper imports for ServerCapabilities and ToolsCapability from mcp.types
- Now correctly declares tools capability using ToolsCapability() instead of empty dict
- This follows the MCP protocol specification for server capability declaration
The previous implementation used {"tools": {}} which was not the correct way
to declare tool capabilities according to the MCP Python SDK. This fix ensures
proper capability negotiation during MCP client connection initialization.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Corrected thinking modes table to show `high` as default for thinkdeep (not `max`)
- Moved the default notation from `max` row to `high` row where it belongs
- This aligns with the actual tool implementation which uses "high" as default
The thinkdeep tool uses "high" (16,384 tokens) as its default thinking mode,
not "max" (32,768 tokens) as previously documented.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Removed all "Triggers:" lines from tool descriptions in README
- The trigger words were redundant since prompt examples already show usage
- Makes documentation cleaner and more focused on actual usage patterns
The prompt examples throughout the documentation already demonstrate
how to invoke each tool, making the trigger word lists unnecessary.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Added dedicated "Web Search Integration" section explaining the new behavior
- Updated all web search capability descriptions to reflect that Gemini recommends searches for Claude
- Added to table of contents for easy navigation
- Explained how the feature works with concrete examples
- Listed benefits of the collaborative approach
- Added note about disabling web search when needed
The documentation now clearly explains that:
- Web search is enabled by default
- Gemini analyzes when searches would help
- Provides specific search recommendations for Claude
- Claude performs the actual searches
- More collaborative and transparent approach
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Updated all references in README.md from old tool names to new ones:
- think_deeper → thinkdeep
- review_code → codereview
- review_changes → precommit
- Fixed syntax error in base.py (unterminated f-string literal)
- Removed trailing whitespace to pass linting
- All tests now pass (96 tests)
- All linting checks pass (ruff, black, isort)
This completes the tool renaming refactor with full documentation updates.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Changed use_websearch default from False to True for all tools
- Redesigned websearch behavior to focus on reasoning rather than direct search
- Tools now analyze when web searches would be helpful and recommend specific searches
- Updated all tool-specific websearch instructions to be more thoughtful
- Gemini will now suggest what Claude should search for and explain why
Key improvements:
- Web search is now enabled by default for better documentation awareness
- Instead of performing searches, Gemini reasons about what searches would help
- Provides specific search recommendations for Claude to execute
- More collaborative approach between Gemini and Claude
This change makes the tools more intelligent about when external information
would be valuable while keeping Claude in control of actual web searches.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Renamed `think_deeper` tool to `thinkdeep` for shorter, cleaner naming
- Updated all imports from ThinkDeeperTool to ThinkDeepTool
- Updated all references from THINK_DEEPER_PROMPT to THINKDEEP_PROMPT
- Updated tool registration in server.py
- Updated all test files to use new naming convention
- Updated README documentation to reflect new tool names
- All functionality remains the same, only naming has changed
This completes the tool renaming refactor for improved clarity and consistency.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Renamed `review_code` tool to `codereview` for better naming convention
- Renamed `review_changes` tool to `precommit` to better reflect its purpose
- Updated all tool descriptions to remove "Triggers:" sections and improve clarity
- Updated all imports and references throughout the codebase
- Renamed test files to match new tool names
- Updated server.py tool registrations
- All existing functionality preserved with improved naming
This refactoring improves code organization and makes tool purposes clearer.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Rename debug_issue.py to debug.py
- Update tool name from 'debug_issue' to 'debug' throughout codebase
- Update all references in server.py, tests, and README
- Keep DebugIssueTool class name for backward compatibility
- All tests pass with the renamed tool
This makes the tool name shorter and more consistent with other
tool names like 'chat' and 'analyze'. The functionality remains
exactly the same.
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
- Emphasize Gemini's 1M token capacity in the description
- Explicitly encourage sharing large log files, stack traces, memory dumps
- Add 'diagnose issue' as a trigger phrase
- Instruct Claude to proactively use this tool for debugging
- Encourage sharing comprehensive file paths rather than snippets
- Make clear that more context leads to better debugging analysis
This ensures Claude understands it should liberally share diagnostic
files and logs when debugging, taking full advantage of Gemini's
massive context window for thorough root cause analysis.
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
- Update 'save this output\!' instruction to mention Claude Code option
- Add direct link to Claude Code configuration section
- Helps users who prefer Claude Code over Claude Desktop
This clarifies that users have two configuration options and can
skip the Claude Desktop step if they're using Claude Code directly.
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
- Add clickable link from 'Bypass MCP's token limits' bullet point
- Links directly to the Working with Large Prompts section
- Makes it easier for users to jump to detailed explanation
This improves navigation and helps users quickly find information
about the large prompt handling feature.
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
- Emphasize that setup-docker-env scripts automatically build Docker image
- Clarify that $GEMINI_API_KEY from environment is automatically used
- Remove manual 'docker build' step from instructions
- Add clear 'To update' instructions - just run setup script again
- Update 'Why Docker' section to reflect easy updates
This makes it clearer that Docker setup is fully automated and users
don't need to run any Docker commands manually. The setup scripts
handle everything including building, cleaning old images, and
configuring the environment.
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
- Update format_response to include clear next steps for Claude
- Claude will now:
1. First understand the context of issues by examining referenced code
2. Present improvement options to the user as a clear list
3. Only implement fixes the user selects
- Ensures user has control over which improvements to implement
- Promotes better understanding of issues before suggesting fixes
This change improves the user experience by making code review fixes
more interactive and user-controlled rather than automatic.
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
- Add use_websearch parameter to base ToolRequest class
- All tools now inherit web search capability automatically
- Enhanced description emphasizes brainstorming and architectural design
- Add get_websearch_instruction() helper method to base class
- Update all tools to include web search in their prompts when enabled
- Update README documentation with web search examples
- Fix linting issues (ruff, black formatting)
Web search is particularly useful for:
- Brainstorming sessions and architectural design discussions
- Exploring industry best practices
- Working with specific frameworks/technologies
- Researching solutions to complex problems
- Getting current documentation and community insights
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
Added check in translate_path_for_environment() to detect and skip
already-translated container paths (those starting with /workspace).
This prevents the function from attempting to translate paths like:
- /workspace/src/main.py -> /inaccessible/outside/mounted/volume/workspace/src/main.py
Now it correctly handles:
- Host path: /Users/.../src/main.py -> /workspace/src/main.py (translation)
- Container path: /workspace/src/main.py -> /workspace/src/main.py (no change)
Added comprehensive test to verify double-translation prevention.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Added explicit translation of files parameter using translate_file_paths()
- Now both path and files parameters are properly translated for Docker environments
- Ensures consistent behavior between dockerized and normal installations
- Updated all references to use translated_files consistently
This fixes the bug where files parameter wasn't being explicitly translated,
even though resolve_and_validate_path was handling it internally. The explicit
translation makes the behavior clear and consistent with path translation.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Updated thinking modes section to emphasize that:
- Claude automatically selects appropriate thinking modes based on task complexity
- Manual override is available but not required in most cases
- Auto-managed mode is recommended for optimal cost/quality balance
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Fixed review_changes tool to properly translate host paths to container paths in Docker
- Prevents "No such file or directory" errors when running in Docker containers
- Added proper error handling with clear messages when paths are inaccessible
refactor: Centralized token limit validation across all tools
- Added _validate_token_limit method to BaseTool to eliminate code duplication
- Reduced ~25 lines of duplicated code across 5 tools (analyze, chat, debug_issue, review_code, think_deeper)
- Maintains exact same error messages and behavior
feat: Enhanced large prompt handling
- Added support for prompts >50K chars by requesting file-based input
- Preserves MCP's ~25K token capacity for responses
- All tools now check prompt size before processing
test: Added comprehensive Docker path integration tests
- Tests for path translation, security validation, and error handling
- Tests for review_changes tool specifically with Docker paths
- Fixed failing think_deeper test (updated default from "max" to "high")
chore: Code quality improvements
- Applied black formatting across all files
- Fixed import sorting with isort
- All tests passing (96 tests)
- Standardized error handling follows MCP TextContent format
The changes ensure consistent behavior across all environments while reducing code duplication and improving maintainability.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
