Loading ToolBench
A simple MCP server that provides basic mathematical operations
This server has critical definition gaps. Both tools lack parameter descriptions entirely (null values in schema), violating the core requirement that every parameter must have a description. Tool descriptions are minimal (10-20 chars), below the 34-char p10 baseline. No output schemas are documented. No error handling guidance. The tools are trivial (add, multiply) but even simple tools require complete definitions for LLM reasoning.
Add two numbers
Multiply two numbers
All parameters lack descriptions (null values in schema). Parameters 'a' and 'b' have no explanation of what they represent or constraints.
Tool descriptions are too short (10-20 chars). 'Add two numbers' and 'Multiply two numbers' lack context on when to use each tool, expected input ranges, or return type documentation.
No output schemas documented. LLMs cannot infer that add() returns an int or what the result represents semantically.
No input validation or error handling. Tools will fail silently or with Python exceptions if given non-integer inputs; no recovery guidance for LLMs.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 28 | 2026-07-28+ | v2 |
No parameter constraints (min/max for integers). LLMs could pass arbitrarily large numbers, causing overflow or performance issues.