Two tools with basic schemas and minimal descriptions. Both tools have input schemas with proper types (number, string enum), but descriptions are extremely short (10-40 chars), falling well below the 50-200 char baseline for LLM-optimized descriptions. No output schemas documented. Error handling is minimal (divide-by-zero check in 'calculate' returns text, not structured error). Tool names lack action verbs ('add' and 'calculate' are weak; 'perform_addition' and 'perform_calculation' would be clearer). Parameters lack detailed descriptions explaining constraints, ranges, or use cases. No pagination, no composition guidance, no recovery hints.

• Expand tool descriptions to 50-200 characters. Example for 'add': 'Performs addition of two numbers. Returns the sum as a text string. Use when you need to add two numeric values together.'
• Add descriptions to all parameters. Example: 'a: The first number to add (any real number). b: The second number to add (any real number).' For 'operation': 'The arithmetic operation to perform: add (sum), subtract (difference), multiply (product), or divide (quotient).'
• Document output schema in tool metadata. Declare that both tools return {content: [{type: 'text', text: string}]} so LLMs know the response structure.
• Improve error handling for divide-by-zero. Return structured error: {content: [{type: 'text', text: 'Error: Division by zero. The divisor (b) must be non-zero.'}]} with clear constraint violation.
• Rename tools to action verbs: 'add' → 'perform_addition', 'calculate' → 'perform_arithmetic_operation'. This makes intent explicit and helps LLMs disambiguate when many tools are available.