Loading ToolBench
A sample MCP server for deployment
Single tool 'question' has critical definition gaps. Tool name lacks action verb (should be 'answer_question'). Description is minimal (18 chars, below 20-char threshold). Input schema present but parameter lacks description. No output schema documented. No error handling guidance. Fails multiple hard scoring rules.
Answers any question.
Tool name 'question' lacks action verb. Should be 'answer_question' to signal intent clearly to LLMs.
Description 'Answers any question.' is 18 characters, below 20-char minimum. Expand to explain WHEN to use, WHAT it returns, and any limitations.
Input parameter 'question' has no description. LLMs cannot infer what format or constraints apply. Add: 'The question to answer (any topic, free-form text)'.
No output schema documented. LLMs cannot plan downstream actions or validate the response structure. Document: returns string with the answer.
No error handling or recovery guidance. If the tool fails or returns unexpected output, the LLM has no guidance on what to do next.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 33 | 2026-07-28+ | v2 |