15 tools with consistent verb-noun naming (get_*, web_search). All have descriptions (avg ~100 chars, baseline 194). Input schemas present for all tools with proper JSON Schema types. However, descriptions are generic and lack LLM-optimization guidance (e.g., 'when to use instead of similar tools'). Parameter descriptions are minimal (1-2 sentences). No output schemas documented. Error handling and recovery guidance absent. No tool annotations (readOnlyHint, destructiveHint). Missing dependency hints and multi-step planning guidance.

• Add output schemas for all 15 tools. Document field names, types, and what each field represents. Example: get_weather should return {temperature: number, conditions: string, wind_speed: number, humidity: number}.
• Enhance descriptions to 50-200 chars with LLM-optimization: state WHAT, WHEN to use, and WHAT IT RETURNS. Example: 'Get current weather for Zurich (temperature, wind, humidity). Use this instead of get_pois for weather-specific queries.'
• Add enum constraints to parameters with known values. E.g., waste_type: enum [kehricht, bio, papier, karton]; category: enum [restaurant, hotel, attraction, museum].
• Document parameter formats and ranges. E.g., zip_code: '5-digit Swiss postal code (e.g., 8000)'; limit: 'integer 1-100 (default 10)'; upcoming_days: 'integer 1-365 (default 30)'.
• Add error handling guidance to tool descriptions. E.g., 'If station not found, try search_stations() with partial name. Returns empty list if no results match.'