Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. generateErd has a reasonable description (71 chars) but lacks parameter descriptions for 'collections' and 'format'. generateQuery has a minimal description (54 chars) and no parameter descriptions at all. Neither tool documents output structure, making it impossible for LLMs to plan downstream operations. No error handling guidance. Schemas are present but incomplete, parameters lack descriptions, which violates the 100% baseline for A+ tools.

• Add descriptions to all parameters. For 'collections': 'Optional array of collection names to include. If omitted, all collections in the database are analyzed.' For 'format': 'Output format: mermaid for diagram syntax, json for structured relationship data.'
• Document output schemas. For generateErd: 'Returns {format: string, diagram: string, collections: [{name, fields, relationships}]}'. For generateQuery: 'Returns {query: string, explanation: string, type: string, isValid: boolean}'.
• Expand generateQuery description to 100+ chars: 'Generate MongoDB queries from natural language descriptions. Supports find, insert, update, delete, and aggregation queries. Returns executable query code with optional explanation.'
• Add error recovery guidance. E.g., 'If database not found, call list_databases() first. If query generation fails, try a simpler description or specify the query type explicitly.'
• Add parameter constraints and examples in descriptions. E.g., 'database: MongoDB database name (1-64 alphanumeric chars, no spaces)' and 'description: Natural language query intent (e.g., "find all users over 30 in Seoul")'