Server has 8 tools with explicit schemas and descriptions, but quality is inconsistent. Tool names follow verb_noun convention (query, list_tables, describe_table). Descriptions are present but generic, most lack context on WHEN to use each tool or what distinguishes them. Parameter descriptions are sparse: 'sql' is described, but 'context' in nl_query has no description. Output schemas are not documented, LLMs cannot predict response structure. Error handling is minimal; no recovery guidance. Security concerns: nl_query accepts arbitrary 'context' object without validation; confirm_ddl token mechanism is undocumented. Pagination is present (page, page_size) but lacks total_count or next_cursor in response. Overall: functional but below production baseline.

• Add output schema documentation to every tool. Example for query: 'Returns {rows: [{col1: string, col2: number}], total_count: number, page: number, page_size: number}'.
• Expand nl_query description: 'Translates natural language to SQL. Requires schema context. Returns: {sql: string, result: object}. Note: LLM integration not yet implemented.'
• Document confirm_ddl token: 'Token is a UUID valid for 5 minutes. Invalid/expired tokens return 400 with message: "Token expired or invalid. Request a new DDL confirmation."'
• Add per-parameter descriptions for all tools. Example: schema parameter should say 'PostgreSQL schema name (default: public). Use list_tables() to discover available schemas.'
• Implement error classification in responses. Return {error: string, code: 'INVALID_SQL'|'PERMISSION_DENIED'|'TIMEOUT', retryable: boolean, suggestion: string}.

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry