Loading ToolBench
A remote MCP server with OAuth 2.0 authentication provider, built on Cloudflare Workers. Demonstrates MCP tool registration and OAuth authorization flows.
Single tool 'add' has a basic schema with typed parameters (a, b as numbers) and descriptions, but falls short of production quality. Description is minimal (18 chars, below 20-char threshold). No output schema documented. No error handling guidance. No parameter constraints (min/max). Tool does one thing but lacks depth expected for agent-ready tools. Naming is generic (verb_noun pattern present but 'add' is overly simple for a demo). No idempotency hints, no recovery guidance on failure.
Adds two numbers together
Description too short (18 chars). 'Adds two numbers together' lacks context for LLM selection, when to use it, and what it returns. Minimum 20 chars; target 50-200 chars.
No output schema documented. LLMs cannot plan downstream calls or extract results without knowing the response structure. Document return type (e.g., {result: number}).
No parameter constraints (min/max, precision). Unbounded numbers let LLMs pass absurd values (e.g., 1e308). Add numeric bounds and precision hints.
No error handling guidance. If addition overflows or inputs are invalid, LLM has no recovery path. Add error classification and actionable recovery hints.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 41 | 2026-07-28+ | v2 |
Tool name 'add' is generic and lacks specificity. Consider 'add_numbers' or 'calculate_sum' to improve clarity and distinguish from other arithmetic tools if they exist.