An MCP server for generating AI-powered educational content, specifically coloring book pages for children
Single tool 'coloring_book_page' has a reasonable name and basic description, but critical gaps undermine quality. The input schema is visible and typed (string), but lacks constraints, examples, or guidance on prompt format. The description (58 chars) is below the 194-char baseline and doesn't explain when/why to use the tool or what output format to expect. No output schema is documented, LLMs cannot plan downstream actions or understand the response structure. No error handling guidance. The tool performs a WRITE operation (image generation) but lacks idempotency markers, confirmation patterns, or rate-limit guidance. Overall, this is a minimal implementation that would not pass production code review.
Generate a coloring book image from a prompt
No output schema documented. LLMs cannot infer response structure (image URL? base64? metadata?), forcing guesswork and downstream failures.
Tool description is 58 characters, below the 194-char baseline. Lacks WHAT (what does it return?), WHEN (when should I call this vs. alternatives?), and prerequisites (API key? rate limits?).
Input parameter 'prompt' has no constraints (length, format, content restrictions). LLMs will pass arbitrary strings, risking API errors or hallucinated outputs.
WRITE operation (image generation) lacks idempotency markers, dry-run support, or confirmation pattern. Agents may retry and generate duplicate images.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 47 | 2026-07-28+ | v2 |
No error handling guidance. If the API fails (rate limit, invalid prompt, service down), LLMs receive no recovery path, no retry guidance, no fallback suggestions.