docs/zh-CN/skills/skill-stocktake/SKILL.md
用于审计Claude技能和命令的质量。支持快速扫描(仅变更技能)和全面盘点模式,采用顺序子代理批量评估。
npx skillsauth add affaan-m/everything-claude-code skill-stocktakeInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
斜杠命令 (/skill-stocktake),用于使用质量检查清单 + AI 整体判断来审核所有 Claude 技能和命令。支持两种模式:用于最近更改技能的快速扫描,以及用于完整审查的全面盘点。
该命令针对以下相对于调用命令所在目录的路径:
| 路径 | 描述 |
|------|-------------|
| ~/.claude/skills/ | 全局技能(所有项目) |
| {cwd}/.claude/skills/ | 项目级技能(如果目录存在) |
在第 1 阶段开始时,该命令会明确列出找到并扫描了哪些路径。
要包含项目级技能,请从该项目根目录运行:
cd ~/path/to/my-project
/skill-stocktake
如果项目没有 .claude/skills/ 目录,则只评估全局技能和命令。
| 模式 | 触发条件 | 持续时间 |
|------|---------|---------|
| 快速扫描 | results.json 存在(默认) | 5–10 分钟 |
| 全面盘点 | results.json 不存在,或 /skill-stocktake full | 20–30 分钟 |
结果缓存: ~/.claude/skills/skill-stocktake/results.json
仅重新评估自上次运行以来发生更改的技能(5–10 分钟)。
~/.claude/skills/skill-stocktake/results.jsonbash ~/.claude/skills/skill-stocktake/scripts/quick-diff.sh \ ~/.claude/skills/skill-stocktake/results.json
(项目目录从 $PWD/.claude/skills 自动检测;仅在需要时显式传递)[]:报告“自上次运行以来无更改。”并停止bash ~/.claude/skills/skill-stocktake/scripts/save-results.sh \ ~/.claude/skills/skill-stocktake/results.json <<< "$EVAL_RESULTS"运行:bash ~/.claude/skills/skill-stocktake/scripts/scan.sh
脚本枚举技能文件,提取 frontmatter,并收集 UTC 修改时间。
项目目录从 $PWD/.claude/skills 自动检测;仅在需要时显式传递。
从脚本输出中呈现扫描摘要和清单表:
扫描中:
✓ ~/.claude/skills/ (17 个文件)
✗ {cwd}/.claude/skills/ (未找到 — 仅限全局技能)
| 技能 | 7天使用 | 30天使用 | 描述 | |-------|--------|---------|-------------|
启动一个 通用代理 工具子代理,并使用完整的清单和检查项:
Agent(
subagent_type="general-purpose",
prompt="
根据检查清单评估以下技能清单。
[INVENTORY]
[CHECKLIST]
为每项技能返回 JSON:
{ \"verdict\": \"Keep\"|\"Improve\"|\"Update\"|\"Retire\"|\"Merge into [X]\", \"reason\": \"...\" }
"
)
子代理读取每项技能,应用检查项,并返回每项技能的 JSON 结果:
{ "verdict": "Keep"|"Improve"|"Update"|"Retire"|"Merge into [X]", "reason": "..." }
分块指导: 每个子代理调用处理约 20 个技能,以保持上下文可管理。在每个块之后将中间结果保存到 results.json (status: "in_progress")。
所有技能评估完成后:设置 status: "completed",进入第 3 阶段。
恢复检测: 如果在启动时找到 status: "in_progress",则从第一个未评估的技能处恢复。
每个技能都根据此检查清单进行评估:
- [ ] 已检查与其他技能的内容重叠情况
- [ ] 已检查与 MEMORY.md / CLAUDE.md 的重叠情况
- [ ] 已验证技术引用的时效性(如果存在工具名称 / CLI 参数 / API,请使用 WebSearch 进行验证)
- [ ] 已考虑使用频率
判定标准:
| 判定 | 含义 | |---------|---------| | Keep | 有用且最新 | | Improve | 值得保留,但需要特定改进 | | Update | 引用的技术已过时(通过 WebSearch 验证) | | Retire | 质量低、陈旧或成本不对称 | | Merge into [X] | 与另一技能有大量重叠;命名合并目标 |
评估是整体 AI 判断 — 不是数字评分标准。指导维度:
原因质量要求 — reason 字段必须是自包含且能支持决策的:
"Superseded""disable-model-invocation: true already set; superseded by continuous-learning-v2 which covers all the same patterns plus confidence scoring. No unique content remains.""Overlaps with X""42-line thin content; Step 4 of chatlog-to-article already covers the same workflow. Integrate the 'article angle' tip as a note in that skill.""Too long""276 lines; Section 'Framework Comparison' (L80–140) duplicates ai-era-architecture-principles; delete it to reach ~150 lines.""Unchanged""mtime updated but content unchanged. Unique Python reference explicitly imported by rules/python/; no overlap found."| 技能 | 7天使用 | 判定 | 原因 | |-------|--------|---------|--------|
~/.claude/skills/skill-stocktake/results.json:
evaluated_at:必须设置为评估完成时的实际 UTC 时间。
通过 Bash 获取:date -u +%Y-%m-%dT%H:%M:%SZ。切勿使用仅日期的近似值,如 T00:00:00Z。
{
"evaluated_at": "2026-02-21T10:00:00Z",
"mode": "full",
"batch_progress": {
"total": 80,
"evaluated": 80,
"status": "completed"
},
"skills": {
"skill-name": {
"path": "~/.claude/skills/skill-name/SKILL.md",
"verdict": "Keep",
"reason": "Concrete, actionable, unique value for X workflow",
"mtime": "2026-01-15T08:30:00Z"
}
}
}
data-ai
Design task-local harnesses, eval gates, and reusable skill extraction for Claude dynamic workflow mode and other adaptive agent harnesses.
development
React component testing with React Testing Library, Vitest/Jest, MSW for network mocking, accessibility assertions with axe, and the decision boundary between component tests and Playwright/Cypress end-to-end runs. Use when writing or fixing tests for React components, hooks, or pages.
tools
React and Next.js performance optimization patterns adapted from Vercel Engineering's React Best Practices (https://github.com/vercel-labs/agent-skills). Organizes 70+ rules across 8 priority categories — waterfalls, bundle size, server-side, client fetching, re-render, rendering, JS micro-perf, advanced. Use when writing, reviewing, or refactoring React/Next.js code for performance.
tools
React 18/19 patterns including hooks discipline, server/client component boundaries, Suspense + error boundaries, form actions, data fetching, state management decision trees, and accessibility-first composition. Use when writing or reviewing React components.