This server has 10 tools with READ_ONLY risk. Overall definition quality is poor: most tools lack comprehensive parameter descriptions, none have documented output schemas, and error handling is minimal. Tools are reasonably named with action verbs (find_, list_, read_, get_), but parameter documentation is sparse. For example, 'path' parameters appear in 6 tools but lack format/constraint descriptions (e.g., 'does tilde expansion work? What about relative paths?'). The 'max_results' parameter in search_files has no bounds specified. Descriptions are present but brief (10-50 chars typically), offering minimal LLM guidance on WHEN to use each tool vs. similar alternatives. No output schemas are documented, forcing LLMs to infer response structures. Error handling is rudimentary, tools return plain exception messages with no recovery guidance. Tools like analyze_chat and find_chats operate on the same domain (WhatsApp) but naming doesn't clearly distinguish their scope.

• Document output schemas for all 10 tools. Example for get_forecast: '{"temperature": number, "condition": string, "humidity": number, "wind_speed": number}'. Use JSON Schema format.
• Expand parameter descriptions to 50 - 150 chars. For 'path' params: specify if tilde-expansion is supported, whether relative paths work, max length (255 chars?), and whether it must exist. Example: 'Directory path to search. Supports ~ for home directory. Relative paths resolved from current working directory. Max 255 chars.'
• Add min/max bounds to numeric parameters. Example: 'max_results (integer, 1 - 1000, default 20)' and 'max_size_mb (integer, 1 - 100, default 10)'.
• Rewrite tool descriptions to clarify scope and differentiation. Example: 'find_chats: List all WhatsApp chat exports found in ~/Desktop/WhatsAppChat or ~/Downloads. Call this first to discover available chats.' vs. 'find_whatsapp_chats: Search for WhatsApp chat exports in a specific directory. Use when you know the target path.'
• Add error recovery guidance. Instead of 'Error analyzing chat: File not found', return 'File not found at '/path/to/file'. Check the path exists and is a valid WhatsApp export (starts with [timestamp]). Use find_chats() to list available exports.'
• Return structured JSON from analyze_chat: '{"total_messages": 542, "participants": [{"name": "Alice", "message_count": 312, "percentage": 57.6}]}' instead of formatted text.
• Add pagination/limits to list_directory and search_files. Cap default results at 20 - 50. Document in descriptions: 'Returns up to 50 results. Use offset/limit parameters for pagination.'