ToolDock provides 6 tools with complete JSON schemas and basic descriptions. All tools have input schemas with typed parameters and descriptions. However, descriptions are minimal (10-20 chars on average), lacking context for LLM selection and missing guidance on when to use each tool vs. alternatives. Parameter descriptions are generic ('city parameter', 'weight parameter') rather than actionable. No output schemas documented. Error handling is absent, tools return raw dicts without recovery guidance. Tool composition is reasonable (single responsibility), but naming could be more explicit (e.g., 'calculate_shipping' vs 'get_shipping_quote'). Security annotations present (risk levels) but no permission gates or audit trails visible.

• Expand parameter descriptions from placeholders to actionable guidance. E.g., 'city (string, 2-50 chars): The destination city for shipping calculation. Examples: Delhi, Mumbai, Bangalore.' Include constraints, formats, and valid ranges.
• Expand tool descriptions to 50-200 chars. Include: what the tool does, when to use it (vs. similar tools), what it returns, and any prerequisites. E.g., 'Calculate shipping cost for a package to a specific city. Use this to get a quote before confirming an order. Returns price in INR. Requires valid city name.'
• Document output schemas for all tools. Specify field names, types, and meanings. E.g., calculate_shipping returns {city: string, price: float}. This enables LLMs to chain tools and extract required data.
• Add error handling with recovery guidance. E.g., if city is invalid, return 'City not found. Try search_cities() to find valid destinations.' If order_id doesn't exist, return 'Order not found. Try find_order() with a different ID or search_orders() to list available orders.'
• Add a confirm_clear_demo_cache tool or dry-run parameter to clear_demo_cache. Agents should preview what will be deleted before executing an irreversible operation.