MCP Weather Server that provides weather information for cities
The Weather MCP server has a single tool with a minimal implementation. While the tool is properly registered using FastMCP's @mcp.tool() decorator with Python type hints, the definition quality is weak across multiple dimensions. The description is adequate at 38 characters but provides no guidance on when to use the tool or what the actual output format is (it returns raw weather data from wttr.in, not a structured response). The input schema is minimal, only a 'city' parameter with basic string type and a short description. No output schema is documented. Error handling is present but generic ('Error fetching weather data' with no guidance on recovery). The tool lacks composition hints (e.g., how results should be parsed, why the format is text-based rather than structured JSON) and does not follow the pattern of declaring what data types users can pass (e.g., 'full city name like "New York" or "London"'). The implementation is functional for a tutorial but falls short of production-grade tool engineering standards.
Get the current weather for a given city
Output schema is not documented. LLMs cannot predict the format of returned data (raw weather string vs structured JSON) or plan downstream processing.
Parameter description is generic and does not specify expected format or examples. 'The name of the city' does not indicate whether to pass 'New York', 'new york', 'NYC', or numeric city codes.
Tool description lacks context on WHEN to use it vs other tools or APIs, and does not explain that results are raw text strings from wttr.in, not structured data.
Error handling is present but not actionable. The generic 'Error fetching weather data' message provides no guidance on what went wrong (network failure, invalid city, rate-limit) or what the LLM should try next.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-21 | F | 9 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 44 | - | v1 |
No pagination or result-limiting logic. If wttr.in returns multi-line weather data, there is no indication of size limits or how to interpret the raw response.