A Discord bot powered by LLMs with MCP tool integration, supporting multiple AI providers (OpenAI, Anthropic, Google, Groq, etc.) and Model Context Protocol servers
This server exposes only 1 tool (getWeather) with minimal definition quality. The tool has a basic description and input schema, but lacks depth, error handling guidance, and output documentation. The schema is present but minimalistic. The description is generic and does not explain when to use the tool, what it returns, or error recovery paths. No parameter validation constraints, no output schema documentation, and no error categorization. The tool name is acceptable (verb_noun form), but the overall definition falls well short of production standards.
Get the weather in a location
No output schema documentation. LLMs cannot predict what fields the tool returns (e.g., temperature, humidity, forecast), forcing them to guess or make assumptions.
Description is generic and too brief (28 chars). Does not explain what data is returned, when to call it vs. other weather services, or how to handle errors (e.g., location not found).
Parameter 'location' lacks validation constraints. No enum, pattern, or format hint. Description does not specify: Is it a city name, ZIP code, coordinates? What format? Does the tool accept partial matches?
No error handling guidance. The tool offers no recovery hints if location is not found, API is unavailable, or parameters are invalid. LLMs will have no idea how to retry or escalate.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 37 | <=2025-11-25 | v2 |
| 2026-03-09 | F | 26 | - | v1 |
Tool definition inferred from file reference (extensions/example.ts) but actual implementation/registration code not visible in provided source. Cannot verify schema completeness or confirm the tool actually works as documented.