Loading ToolBench
Remote MCP server with OAuth authentication demo running on Cloudflare Workers
Single tool 'add' has a clear verb-noun name and basic schema with typed parameters. However, the description is minimal (22 chars), lacks context on when/why to use it, and provides no guidance on output format or error handling. Parameters have type definitions but descriptions are generic. No output schema documented. No error handling or recovery guidance. Lacks composition context, unclear how this tool fits into a larger workflow. Baseline: production tools average 194 chars for descriptions and 72 chars per param; this server falls well short.
Adds two numbers together
Tool description is too brief (22 chars) and lacks actionable context. Does not explain when to use this tool, what it returns, or error conditions.
No output schema documented. LLM cannot infer what fields the response contains or plan downstream operations.
Parameter descriptions are generic ('First number to add', 'Second number to add'). No constraints on range, precision, or edge cases (e.g., can inputs be negative? What about very large numbers?).
No error handling guidance. What happens if inputs are NaN, Infinity, or non-numeric? How should the LLM recover?
Tool serves no clear business purpose in isolation. Lacks context on composition, is this a demo/test tool, or part of a calculator workflow? Ambiguous intent reduces LLM confidence in selection.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |