Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output clarity. query_table has a proper Pydantic output model (QueryResult), but analyze_table returns a bare dict with no schema. Parameter descriptions are minimal (10-20 chars). No error handling guidance, no enum constraints, no validation rules stated. Naming is acceptable (verb_noun pattern), but descriptions lack context on WHEN to use each tool or what distinguishes them. Missing: per-parameter descriptions for analyze_table, output schema documentation for analyze_table, error recovery guidance, and actionable error messages.

• Add explicit output schema to analyze_table: define a Pydantic model (e.g., AnalysisResult) with fields like row_count, column_stats, data_types, and return it instead of dict.
• Expand parameter descriptions: 'limit: Max rows to return (1 - 1000, default 10). Larger limits may slow queries.' This sets bounds and explains trade-offs.
• Add WHEN-to-use context to tool descriptions: 'query_table: Fetch raw data rows. Use when you need specific records. analyze_table: Get summary statistics (count, types, nulls). Use to understand table structure before querying.'
• Implement error recovery guidance: catch ValueError for missing tables and return 'Table "X" not found. Available tables: users, products, orders. Did you mean one of these?'
• Add enum constraint for table_name parameter: define allowed values (users, products, orders) as an enum in the schema to prevent hallucinated table names.