MCP server for submitting restaurant feedback with text and optional base64-encoded images to MongoDB
Single tool with basic schema and minimal descriptions. SubmitFeedback has a complete JSON Schema with typed parameters (string, number, array), but descriptions are extremely terse (5-30 chars), falling well below the 50-200 char baseline for LLM-optimized descriptions. No error handling guidance, no output schema documentation, no parameter constraints (e.g., rating range, email validation beyond format), and no indication of side effects (data persistence). The tool name lacks a clear action verb prefix (should be 'submit_feedback' per convention). Missing idempotency hints, retry guidance, and recovery instructions.
Submit a feedback
Tool description is 18 characters ('Submit a feedback'), far below 50-200 char baseline. LLMs cannot determine when/why to select this tool or what it returns.
Parameter descriptions are trivial (5-30 chars: 'Customer name', 'Customer email', 'Restaurant rating', 'Restaurant comment'). No format constraints, ranges, or usage guidance. LLMs cannot infer valid input bounds.
No output schema documented. Tool returns a text response with submission ID, but LLMs have no structured schema to parse or chain downstream calls. Missing documented return type.
No error handling or recovery guidance. If validation fails (invalid email, missing fields, image decode error), the handler throws generic errors with no actionable next steps for the LLM.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 49 | 2026-07-28+ | v2 |
Tool name 'SubmitFeedback' does not follow verb_noun convention (should be 'submit_feedback'). Inconsistent with 90% of production tools that start with action verbs in lowercase.