Server has 9 tools with mixed quality. Most tools have descriptions (7/9), but parameter documentation is sparse and inconsistent. Schemas are present but minimal, many parameters lack type information or descriptions. Tool naming is reasonable (verb_noun pattern mostly followed), but several tools lack clarity on when to use them vs. alternatives. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite all tools being read-only. Error handling guidance is absent. Output schemas are not documented. This is a typical community-grade server with foundational structure but significant gaps in LLM-optimization.
Create vector embedding for text
Generate text using LLM
Returns a bounded app/package routing tree for the current project.
Read one Nebutra UI component contract by registry id.
Read dry-run codemod and migration hints for one component.
Search the Nebutra UI agent manifest by query, tag, status, or maturity.
Validate production evidence for one component or the full UI contract.
No tool annotations despite all tools being read-only. Missing readOnlyHint, destructiveHint, and idempotentHint metadata that LLMs use to reason about safety and retry logic.
Output schemas not documented. LLMs cannot plan downstream tool calls or extract required fields without knowing what each tool returns. E.g., nebutra_ui_search_components returns what fields? Does it include component_id, status, maturity, or all three?
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 50 | 2026-07-28+ | v2 |
Return pong
Translate text between languages
Parameter descriptions are minimal or missing context. E.g., 'root' and 'manifestPath' in nebutra_ui_* tools lack explanation of when/why to override defaults. 'temperature' in generate_text has no range guidance (0.0 - 2.0?). LLMs cannot infer valid ranges or use cases.
ping tool is deprecated in MCP spec (removed as of 2026-07-28). Should be removed or replaced with a health-check resource. Keeping it signals outdated implementation.
No error handling guidance. Tools lack descriptions of failure modes, recovery steps, or when to retry. E.g., what happens if generate_text hits a rate limit or API timeout? How should the LLM respond?