MCP server that provides weather information for cities using OpenWeatherMap API
Single tool with critical definition gaps. Tool name 'get_weather' follows verb_noun convention (+), but description is non-English ('Kisi bhi city ka weather batao' = Hindi for 'Tell me weather of any city'), creating LLM comprehension risk. Input schema present with Zod validation, but parameter description is also in Hindi. No output schema documented. Error handling returns raw API errors without recovery guidance. Tool lacks idempotency hints, rate-limit guidance, and structured output documentation. Baseline: 194 chars avg description (this is ~30 chars in English equivalent); 72 chars avg param annotation (this is ~40 chars). No tool annotations (readOnlyHint, etc.) despite READ_ONLY risk classification.
Kisi bhi city ka weather batao
Tool and parameter descriptions in Hindi, not English. LLMs trained primarily on English will misinterpret intent and fail to select tool appropriately.
No output schema documented. Callers cannot predict response structure (fields, types, pagination). Baseline: 100% of A+ tools document return types.
Error handling returns raw API error messages ('Error: {data.message}') without recovery guidance. LLM cannot determine if error is retryable, user-fixable, or fatal.
No tool annotations despite READ_ONLY risk classification. Tool should declare readOnlyHint=true to signal safety to agents.
API key exposed via environment variable without validation. No rate-limit handling or timeout configuration for external API calls.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 39 | 2026-07-28+ | v2 |