Two well-defined tools with complete schemas, clear descriptions, and proper parameter typing. Both tools have enums and constraints (calculate: operation enum, find_catalog: limit bounds 1-20). Descriptions are concise (50-80 chars) and action-oriented. Tool names follow verb_noun pattern (calculate, find_catalog). Security is enforced via require_scope() calls. Main gaps: no error recovery guidance in descriptions, no output schema documentation in docstrings, and missing parameter descriptions in the actual tool signatures (though Pydantic Field provides some via defaults).

• Add output schema documentation to tool docstrings: 'Returns: {operation: str, value: float}' for calculate; 'Returns: list of {id, name, category, description}' for find_catalog.
• Enhance error descriptions with recovery paths: 'Raises ValueError if right=0 (divide-by-zero). Caller should validate operands or retry with non-zero divisor.'
• Add parameter descriptions via docstring or Pydantic Field descriptions: operation='Arithmetic operation: add, subtract, multiply, or divide', left='Left operand (number)', right='Right operand (number)'.
• Declare tool annotations in MCPServer registration: mark both tools with readOnlyHint=true and idempotentHint=true to signal safety to clients.
• Document the scope requirements in tool descriptions: 'Requires scope: calculator:use' and 'Requires scope: catalog:read' so agents know what permissions are needed.

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

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