Ada Diamonds MCP server has 6 tools with complete input schemas and descriptions. Naming follows verb_noun convention (search, list, read). Descriptions are present but brief (10-50 chars), below the 50-200 char LLM-optimized baseline. All tools are read-only (no state mutation). Schemas include proper types and parameter descriptions. However, output schemas are not documented, LLMs cannot predict response structure for chaining. No error handling guidance, no pagination details despite search tools returning lists. Parameter constraints (enums for shape, color, clarity, cut) are present but not exhaustive.

• Expand tool descriptions to 80-150 chars. Example: 'diamonds' → 'Search lab-grown loose diamonds by shape, carat, color, clarity, and price. Returns inventory with GIA grades and pricing. Use before rings to find center stones.'
• Document output schemas for all tools. Specify fields returned (e.g., diamonds returns [{ id, shape, carat, color, clarity, cut, price, url }]), types, and which fields are safe to pass to downstream tools.
• Add pagination documentation: 'Results capped at 200. Use limit and offset for pagination. Returns total_count so LLM knows if more results exist.'
• For showrooms, document what it returns: 'Returns array of showroom locations with address, hours, phone, and remote consultation availability.'
• Add error guidance: 'If shape is invalid, returns 400 with available shapes. If price range is empty, returns 200 with empty array, try widening min_price or max_price.'