This is a learning/practice repository with significant quality gaps. Of 5 tools, all have basic descriptions (20-100 chars), but critical issues emerge: (1) schemas are inferred from fragments rather than explicitly visible in tool registration code; (2) descriptions lack context on WHEN to use tools and prerequisites; (3) no parameter descriptions beyond name; (4) no error handling guidance; (5) no output schema documentation; (6) missing tool composition planning (e.g., search_papers stores data but extract_info retrieves it, relationship unexplained). The codebase shows client examples (mcp_chatbot_for_multi_server.py) but the actual server implementations (chatbot_example.py, mcp_server_by_youtube/main.py) are not fully included, only fragments are visible. This means tool definitions cannot be fully audited.

• Add explicit, detailed parameter descriptions to every tool. For employee_id, specify: 'The unique employee identifier (numeric ID, e.g. 12345, or email, e.g. john@example.com).' For leave_dates, add: 'Array of dates in YYYY-MM-DD format. Cannot be past dates or more than 30 consecutive days.'
• Document output schemas for all tools. E.g., search_papers should return: [{paper_id: string, title: string, authors: string[], published_date: string}]. get_leave_balance should return: {employee_id: string, total_days: integer, used_days: integer, remaining_days: integer}.
• Clarify tool composition. Add to search_papers: 'Call this first to discover papers. Use extract_info to retrieve full details of a known paper_id.' Add to extract_info: 'Requires a paper_id from search_papers or external knowledge.'
• Add error recovery guidance to apply_leave. E.g., 'If request fails due to insufficient balance, call get_leave_balance first to show available days. If dates conflict, call get_leave_history to show existing approvals.'
• Add constraints. search_papers max_results: 'Default 5, maximum 50. Higher values may timeout.' apply_leave leave_dates: 'Cannot exceed 30 consecutive days or be in the past.'
• For apply_leave, implement a confirmation pattern: add a boolean 'confirm' parameter defaulting to false. When false, return a preview. When true, execute.
• Provide all source code. Currently chatbot_example.py and mcp_server_by_youtube/main.py are referenced but not shown. Include full tool registration with @server.tool() decorators, input schemas, and implementation logic.