A remote MCP server example deployed on Cloudflare Workers with OAuth 2.0 authentication, demonstrating MCP tools integrated with external APIs like PokeAPI
Server has 2 tools with minimal descriptive quality and incomplete schemas. The 'add' tool lacks any description at all (scoring 0 for description). The 'getPokemonInfo' tool has a description but parameters lack descriptions, only the 'pokemon' parameter has a description. Both tools are missing documented return schemas and error handling guidance. Parameters lack type constraints, ranges, and format documentation. No input validation is visible in tool definitions, and error responses don't guide LLM recovery. The server shows basic tool registration via workers-mcp framework, but definitions fall well short of production quality baseline.
Get information about a Pokemon from PokeAPI, including abilities and their effects.
Tool 'add' completely lacks a description. LLMs cannot determine when to invoke it or why it exists.
Parameters in 'add' tool lack descriptions entirely. The parameters 'a' and 'b' have no explanation of what they represent. LLMs must guess based on name alone.
No documented output schemas visible for either tool. LLMs do not know what fields to expect in responses, making it impossible to plan downstream tool calls or extract data reliably.
Error handling is minimal and generic. The getPokemonInfo tool catches errors and returns text, but error messages do not guide the LLM on recovery actions or classify errors as retryable/fatal.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 5 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 39 | - | v1 |
No input validation or constraints visible. The 'pokemon' parameter accepts any string, no guidance on format, length, or valid values. LLMs may pass malformed input.
The 'add' tool name is generic. Verb_noun naming (e.g., 'compute_sum' or 'calculate_total') would better convey the tool's intent and distinguish it in a larger tool set.