Local MCP server with tools for greeting and suggesting new MCP ideas
Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output specification. Tool 1 (hello-world) has minimal description (24 chars, below baseline 194 avg). Tool 2 (new-idea-for-mcp) has verbose description (200+ chars) but lacks clarity on what 'successfully created' means in the response. Neither tool documents output schema structure. Parameter descriptions are present but sparse, 'name' parameter has null description in Tool 1. No error handling guidance, no idempotent hints, no output field documentation. Naming is acceptable (verb-noun style) but Tool 2 name is awkwardly long and compound ('new-idea-for-mcp' suggests multiple concerns). No pagination, no structured output examples, no recovery guidance.
Say hello to someone
Suggest a new idea for MCP.AI Agent is developing, and if you have an idea for a new feature for MCP, please suggest it.This aims to accelerate development by the AI Agent.Execution timing is before the task is completed, if you have an idea.
Tool 1 'hello-world': parameter 'name' has null description. All parameters must have non-empty descriptions explaining what they control.
No output schema documented for either tool. LLMs cannot plan downstream calls or extract data without knowing response structure (fields, types, cardinality).
Tool 1 description is 24 characters ('Say hello to someone'), below 34-char p10 baseline. Lacks context on when/why to use it vs alternatives.
Tool 2 'new-idea-for-mcp' name is compound and vague. Does it suggest, create, submit, or track ideas? Rename to 'create_github_issue' or 'submit_feature_idea' for clarity.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 39 | 2026-07-28+ | v2 |
No error handling guidance. If GitHub API fails (auth, rate limit, network), handlers return raw errors. LLM cannot determine if retry is safe or what to do next.