Two tools with basic schemas and minimal descriptions. Both tools have input schemas with proper types (number, string, enum), but descriptions are extremely short (10-50 chars) and lack context for LLM selection. No output schema documentation. No error handling guidance beyond a single divide-by-zero check. Tool names lack action verbs (add, calculate are weak; should be compute_sum, perform_arithmetic). Parameters lack descriptions entirely in the schema, only visible in code comments. No pagination, no structured output guidance, no recovery hints.

• Expand tool descriptions to 50-200 chars. Example: 'add: Computes the sum of two numbers. Use when you need basic arithmetic addition. Returns the numeric result.' This matches the 194-char baseline for production tools.
• Add descriptions to all parameters in the Zod schema. Example: z.number().describe('First number to add (any real number)'). This ensures LLMs see constraints and context.
• Document output schema explicitly. Example: 'Returns: { result: number, operation: string, timestamp: ISO8601 }'. This enables LLMs to chain tools and validate results.
• Rename tools with clearer action verbs. 'add' → 'compute_sum', 'calculate' → 'perform_arithmetic'. This helps LLMs disambiguate when many tools are available.
• Add comprehensive error handling with recovery guidance. Example: 'Invalid operation: got "modulo", must be one of: add, subtract, multiply, divide. Try again with a valid operation.' This enables LLM self-correction.