Loading ToolBench
A simple MCP server providing calculator tools without authentication
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 schema documentation. 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). No parameter descriptions beyond type hints. Schemas are present but sparse.
Simple addition tool
Calculator tool with multiple operations
Tool descriptions are critically short (10-40 chars). 'Simple addition tool' and 'Calculator tool with multiple operations' lack context for LLM selection. Should explain WHAT the tool does, WHEN to use it, and what it returns (e.g., 'Adds two numbers and returns their sum as a numeric result. Use when you need basic arithmetic addition.').
Parameter descriptions are missing entirely. Zod schema defines types (number, enum) but LLMs cannot infer meaning from names alone. 'a' and 'b' are ambiguous, are they operands, indices, or something else? Add descriptions like 'First operand (numeric value)' and 'Second operand (numeric value)'.
Output schema is not documented. Callers cannot predict the response structure. Document that both tools return {content: [{type: 'text', text: string}]} and explain what the text field contains (numeric result or error message).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | 2026-07-28+ | v2 |
Tool names lack action verbs. 'add' and 'calculate' are weak; they do not clearly distinguish intent. Rename to 'perform_addition' or 'add_numbers' and 'perform_arithmetic' or 'calculate_expression' to match verb_noun convention and improve LLM disambiguation.
Error handling is unstructured. The 'calculate' tool returns a text error message for division by zero ('Error: Cannot divide by zero') instead of a structured error response. Per pattern:recovery-guide, errors should tell the LLM what to do next and be categorized as retryable or fatal.