Server has 6 tools with complete input schemas and descriptions, but quality is inconsistent. All tools follow verb_noun naming (fortune_*), but descriptions are domain-specific Chinese fortune-telling terminology that may not clearly convey to LLMs when/why to use each tool. No output schemas documented, tools return {htmlPath, summary, html} but this is not declared in tool definitions. Parameters lack validation constraints (e.g., year range, month 1-12 bounds not enforced in schema). Error handling is minimal, generic catch-all returns 'isError: true' with message only, no recovery guidance. No tool annotations (readOnlyHint, etc.). Composition is reasonable (each tool does one thing), but lack of structured output documentation and missing parameter constraints significantly limit LLM usability.
八字排盘 — 输入出生年月日时和性别,推算四柱八字、五行分布、日主分析和今日运势
程序员黄历 — 输入编程语言,看今天的宜忌、代码气运和幸运语言
六爻起卦 — 心中默念问题,自动起卦得本卦变卦,解读卦象
起名 — 输入姓氏和出生信息,根据八字五行补缺生成候选名字
姓名打分 — 输入姓名,五格剖象法分析天格/人格/地格/外格/总格,综合评分
抽签 — 从灵签筒中抽取一支签,得签文和解签白话文
No output schema documented. Tools return {htmlPath, summary, html} but this structure is not declared in tool definitions. LLMs cannot plan downstream actions or extract structured data.
Parameter constraints missing from schemas. fortune_bazi accepts year/month/day/hour but no min/max bounds declared (e.g., month should be 1-12, hour 0-23). LLMs can pass invalid values.
Error handling is generic. Catch-all returns 'isError: true' with message only. No recovery guidance (e.g., 'Invalid year: must be 1900-2100') or error classification (retryable vs fatal).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 50 | 2026-07-28+ | v2 |
Descriptions use domain-specific Chinese terminology (八字, 五行, 六爻) without explaining to LLMs when/why to call each tool. fortune_bazi vs fortune_mingming distinction unclear from descriptions alone.
No tool annotations. Tools are read-only (no state mutation) but lack readOnlyHint annotation. Agents cannot infer safety/idempotency from definitions.