AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Skill Auditor

skill-zlhad-agent-skills-skill-auditor · by ZLHad

>

No reviews yet
0 installs
12 views
0.0% view→install

Install

$ agentstack add skill-zlhad-agent-skills-skill-auditor

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-zlhad-agent-skills-skill-auditor)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
3mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Skill Auditor? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Skill Auditor — 自定义 Skill 质量审查

审查当前项目中所有自定义 Claude Code Skills,确保符合渐进式披露规范和最佳实践。

审查维度

1. Frontmatter 质量(YAML 头部)

| 检查项 | 规范 | 严重度 | |--------|------|--------| | name | 必填,简洁 | 严重 | | description | 必填,含触发词,50-200 字 | 严重 | | allowed-tools | 按需列出,不多不少 | 警告 | | model | 按复杂度选择(sonnet/opus) | 建议 | | 触发词覆盖 | description 中包含中英文触发短语 | 警告 |

Description 要点:Stage 1 只加载 frontmatter(~50 tokens/skill),description 是触发匹配的唯一依据,必须包含用户可能说的关键词和短语。

2. SKILL.md 正文

| 检查项 | 规范 | 严重度 | |--------|------|--------| | 总行数 | ≤500 行 | 警告 | | 核心工作流 | 必须清晰描述阶段/步骤 | 严重 | | 参考文件链接 | references/ 下的文件必须有 markdown 链接 | 警告 | | 冗余内容 | 大段模板/示例/列表应拆到 references | 建议 | | 术语一致性 | 文件内术语前后一致 | 建议 |

3. 目录结构

skill-name/
├── SKILL.md              <- 必须
├── README.md             <- 建议(公开发布时)
├── references/           <- 建议(补充文档)
│   ├── *.md
├── scripts/              <- 可选(辅助脚本)
├── templates/            <- 可选(输出模板)
└── config/               <- 可选(配置文件)

4. 参考文件完整性

  • references/ 下的 .md 文件是否都被 SKILL.md 链接
  • SKILL.md 中的 references/ 路径是否都指向存在的文件
  • 参考文件行数是否合理(单个建议 ≤200 行)

工作流程

步骤 1:扫描 Skills 目录

扫描目标路径下所有包含 SKILL.md 的子目录,识别为 skill。

默认扫描路径:当前工作目录。

步骤 2:逐个审查

对每个 skill 执行以下检查:

1. 读取 SKILL.md
2. 解析 frontmatter(name, description, allowed-tools, model)
3. 统计正文行数
4. 检查是否有 references/ 目录
5. 如有 references/,检查双向链接完整性:
   - SKILL.md → references/*.md(链接目标是否存在)
   - references/*.md → SKILL.md(文件是否被引用)
6. 检查 README.md 是否存在
7. 汇总问题清单

步骤 3:生成审查报告

按以下格式输出:

# Skill 审查报告

## 总览

| Skill | 行数 | Frontmatter | 参考链接 | README | 总评 |
|-------|------|-------------|----------|--------|------|
| literature-search | 168 | ✅ | ✅ | ✅ | 通过 |
| knowledge-base-manager | 306 | ✅ | ✅ | ❌ | 警告 |

## 详细问题

### skill-name
- ⚠️ [警告] 缺少 README.md
- 💡 [建议] description 可增加英文触发词

严重度图标:

  • ❌ 严重:影响 skill 加载或功能
  • ⚠️ 警告:不规范但不影响功能
  • 💡 建议:可优化项

步骤 4:修复建议

对每个问题给出具体修复建议。如用户确认,可直接执行修复。

渐进式披露加载原理速查

供审查时参考的 Claude Code skill 加载机制:

| 阶段 | 触发时机 | 加载内容 | Token 开销 | |------|----------|----------|------------| | Stage 1 | 会话启动 | 仅 YAML frontmatter | ~50/skill | | Stage 2 | 用户输入匹配 description | 完整 SKILL.md | 1000-5000 | | Stage 3 | Skill 执行中按需读取 | references/ scripts/ | 按需 |

核心原则:Stage 2 加载的 SKILL.md 是 Claude 理解任务的主要依据,必须自包含核心工作流;大段辅助内容放 Stage 3 按需加载。

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.