This server has significant definition quality gaps. Tool descriptions are present but minimal (10-30 chars), falling well below the 50-200 char baseline for LLM-optimized descriptions. Input schemas are visible and typed, but parameter descriptions are sparse or missing context. No output schemas are documented. Error handling is absent, tools provide no recovery guidance. The server lacks idempotency hints, permission gates, and audit trail patterns. Tool naming follows verb_noun convention (get_, mark_, call_), which is good, but descriptions do not explain WHEN to use each tool or what distinguishes similar tools (e.g., get_status vs get_next_key). No tool annotations (readOnlyHint, destructiveHint) are present despite clear WRITE vs READ_ONLY risk classifications.
调用LLM服务(分层轮询策略)
获取全局配置
获取下一个可用的API Key(轮询策略)
获取指定平台的配置
获取所有平台名称
获取Key池状态
加载配置文件
标记Key使用失败
Descriptions are too short (10-30 chars) and lack context. LLMs cannot determine WHEN to call get_status vs get_next_key, or why mark_key_success differs from mark_key_error. Baseline is 50-200 chars with WHAT, WHEN, and dependencies.
No output schemas documented. Callers cannot infer what fields to expect from get_status, get_provider_config, or call_llm responses. This forces LLMs to guess and wastes tokens on exploratory calls.
No error handling or recovery guidance. Tools provide no actionable error messages. If call_llm fails, the LLM has no guidance on whether to retry, call get_next_key, or ask the user.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 44 | 2026-07-28+ | v2 |
标记Key使用成功
验证配置文件格式
Tool annotations missing. mark_key_success and mark_key_error are marked WRITE in the risk field but have no destructiveHint or idempotentHint annotations. LLMs cannot determine if these are safe to retry.
Parameter descriptions are minimal or absent. 'headers' in mark_key_success says 'HTTP响应头(用于提取配额信息)' but does not explain format, required fields, or how quota info is extracted. LLMs cannot construct valid calls.