Three tools with minimal documentation. All have basic descriptions (10-30 chars, below the 50-200 char LLM-optimized baseline), no parameter descriptions, and no documented output schemas. Tool names lack action verbs (itinerary, captions, reels are nouns, not verb_noun patterns). Parameters are typed but undescribed. No error handling guidance, no pagination support, no output structure documentation. This server falls well below production baseline.

• Rename tools to verb_noun pattern: 'generate_itinerary', 'create_captions', 'suggest_reels'. This enables LLMs to parse intent from the name alone.
• Expand tool descriptions to 50-200 chars. Example: 'Generate a detailed travel itinerary with daily activities, dining recommendations, and estimated costs for a specified destination and duration.'
• Add descriptions to all parameters. Example for 'destination': 'The travel destination (city or country name, e.g., Paris, Japan). Used to customize itinerary, captions, and reel ideas.' Example for 'days': 'Number of days for the itinerary (1-30). Determines activity density and scope.'
• Document output schemas. Specify what fields each tool returns (e.g., itinerary returns {day: int, activities: string[], restaurants: string[], estimated_cost: float}). This enables agents to chain tools and extract data.
• Add error handling guidance. Specify which errors are retryable (network timeout) vs user-fixable (invalid destination) vs fatal. Example: 'If destination is not recognized, try a broader region name or call with a nearby major city.'