Server has 5 tools with complete schemas and descriptions. All tools follow verb_noun naming (get-, lookup-, find-, calculate-, check-). Descriptions are present and adequate (60-90 chars), meeting the 10-1024 char baseline. All parameters have type definitions and descriptions. However, descriptions lack LLM-optimization guidance (e.g., 'when to use this tool', 'what it returns'), and output schemas are not documented. Error handling is minimal, tools return generic 'not found' messages without recovery guidance. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite all being read-only operations. Parameter constraints are well-defined (enums, min/max), but descriptions do not explain format expectations or dependencies.

• Add structured output schemas to all tools. Document return fields with types (e.g., 'Returns object with fields: recommendations (array of {title, description, link}), total_count (number)').
• Enhance error messages with recovery guidance. E.g., 'No medication found for "aspirin". Try: search_medications("aspirin") or check_symptoms() to find related conditions.'
• Add readOnlyHint: true to all tool definitions in server.tool() calls to signal safety and enable agent optimizations.
• Expand descriptions to 100-150 chars with explicit 'WHEN' and 'WHAT' guidance. E.g., 'Get preventive care recommendations for a person. Use this to find age/sex-appropriate screenings. Returns list of recommended tests and preventive actions.'
• Document parameter format expectations in descriptions. E.g., 'symptoms: array of symptom names (e.g., ["fever", "cough", "fatigue"]). Each item is a string, 2-50 chars.'