All three tools lack descriptions entirely (0/3 have non-empty descriptions). Parameter descriptions are null for all inputs. No output schemas are documented. Tool names use verb_noun pattern (add_, substract_, multiply_) but descriptions are completely absent, violating the critical requirement that every tool must have a non-empty description for LLM selection. This server fails baseline quality checks across naming, descriptions, and schemas.

• Add a docstring to each tool describing what it does, when to use it, and what it returns. Example: 'Adds two integers and returns their sum. Use this when you need to perform arithmetic addition.'
• Add descriptions to all parameters. Example: 'x: The first integer operand' and 'y: The second integer operand'.
• Document the output schema for each tool. Example: 'Returns an integer representing the result of the operation.'
• Fix the spelling error: rename 'substract_numbers' to 'subtract_numbers' for clarity and correctness.
• Add input validation and error handling with actionable messages. Example: 'Invalid input: x and y must be integers. Got x={type(x).__name__}, y={type(y).__name__}'.

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

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