An MCP server that provides access to Groq's LLM API for chat completions with support for conversation history and model listing
Three tools with basic schemas and Spanish descriptions. All tools have input schemas with type definitions, but descriptions are minimal (10-60 chars, well below the 194-char baseline). Parameters lack depth: 'messages' in chat_with_history has no format specification; 'temperature' lacks range validation in description; 'model' lacks enum constraint despite known fixed set. No output schema documentation. Error handling returns raw exception strings instead of actionable guidance. No tool annotations (readOnlyHint, etc.). Composition is reasonable (three distinct concerns), but parameter design forces LLM guessing.
Envía un mensaje a Groq y devuelve la respuesta.
Envía una conversación completa con historial a Groq.
Lista los modelos de Groq disponibles.
Parameter descriptions are too brief and lack constraints. 'model' parameter accepts only 5 known values but is defined as free-form string with no enum. 'temperature' lacks range specification in description (0.0 - 2.0 mentioned in param description but not enforced). 'messages' format is undocumented.
No output schema documentation. Tools return raw strings or JSON without specifying structure. LLMs cannot plan downstream calls or extract fields reliably. chat_with_history returns a single string; list_models returns JSON but structure is not declared in tool definition.
Error handling returns raw exception strings ('Error: {type}: {message}') instead of actionable guidance. LLM cannot determine if error is retryable, user-fixable, or fatal. No recovery hints provided.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 47 | 2026-07-28+ | v2 |
Descriptions are 10 - 60 characters, well below 194-char baseline. Too brief to guide LLM selection. 'Envía un mensaje a Groq y devuelve la respuesta' (Spanish) does not explain WHEN to use chat vs chat_with_history, or what the return value contains.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). All three tools are read-only but this is not declared in schema, forcing LLM to infer safety from description alone.