Two tools with minimal but present schemas and descriptions. Both tools have basic verb-noun naming (scan, scan_path) and descriptions in Chinese. However, descriptions are terse (under 100 chars), lack WHEN/WHY context, and parameter descriptions are minimal. Input schemas exist but lack detail (no constraints, ranges, or validation hints). No output schema documented. No error handling guidance. Tools are READ_ONLY but lack explicit idempotentHint annotations. No composition guidance or chaining IDs returned. Baseline: 549 production tools average 194 chars for descriptions; these are ~60 chars. Parameter descriptions average 72 chars; these lack detail.
扫描本机默认位置(常见 MCP 配置 + skill 目录),返回安全报告。
扫描指定目录或文件,返回安全报告。
Descriptions lack WHEN/WHY context and are under 100 chars. Rubric baseline: 194 chars average. Current: ~60 chars. LLMs cannot determine when to select scan vs scan_path without explicit guidance.
No output schema documented. Rubric requires: 'Document the output schema. LLMs need to know what fields to expect.' Current code returns JSON with 'ok', 'targets', 'report' but this is not declared in tool definitions.
Parameter descriptions are minimal. 'default_paths' has 'description: 是否扫描默认位置' (5 words); 'path' has 'description: 要扫描的目录或文件路径' (6 words). Rubric baseline: 72 chars average. No format hints, constraints, or examples.
No error handling guidance. Rubric requires: 'Error responses must tell the LLM what to do next.' No documented error cases, recovery paths, or actionable error messages.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 52 | 2026-07-28+ | v2 |
No tool annotations (idempotentHint, readOnlyHint). Both tools are READ_ONLY and idempotent but lack explicit annotations. Rubric: 'tool annotations (readOnlyHint/destructiveHint/idempotentHint)' are current patterns.