Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问clear审计未展示

ln-762-dependency-auditln 762 依赖性审计

Agent Skill

用于辅助安全审计、权限检查、凭据风险、认证流程和常见漏洞排查。它适合让 Agent 梳理敏感配置、检查依赖风险、分析鉴权逻辑或生成安全复核清单。使用时不能把工具输出直接当最终结论,涉及密钥、令牌、用户数据或生产系统时,应先确认最小权限、脱敏方式和操作边界。

总安装

347

周安装

12

GitHub Stars

441

下载量

97
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:ln-762-dependency-audit(ln 762 依赖性审计)
来源仓库:https://github.com/levnikolaevich/claude-code-skills
仓库路径:skills/ln-762-dependency-audit
安装命令:
npx skills add https://github.com/levnikolaevich/claude-code-skills --skill ln-762-dependency-audit
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/levnikolaevich/claude-code-skills --skill ln-762-dependency-audit

简介

用于辅助安全审计和依赖风险检查。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

  • 适合梳理敏感配置、分析鉴权逻辑或生成安全复核清单。
  • 不能将工具输出直接当作最终结论。
  • 涉及密钥或用户数据时应确认最小权限。
  • 需先脱敏再处理敏感信息。ln-762-dependency-audit 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Dependency Audit

Audits project dependencies for known security vulnerabilities across multiple package ecosystems.

Purpose & Scope

  • Detect vulnerable dependencies using ecosystem-specific tools
  • Support multiple ecosystems: npm, NuGet, pip, Go modules, Bundler, Cargo
  • Classify vulnerabilities by severity (Critical/High/Medium/Low)
  • Provide fix recommendations with safe auto-fix guidance
  • Return normalized report to parent orchestrator (ln-760)

When to Use

  • During project bootstrap (via ln-760-security-setup)
  • CI/CD pipeline security checks
  • Pre-release security validation
  • Regular scheduled audits

Workflow

Phase 1: Ecosystem Detection

Step 1: Detect Package Managers

  • Check for package.json / package-lock.json (npm)
  • Check for *.csproj / packages.config (.NET)
  • Check for requirements.txt / Pipfile / pyproject.toml (Python)
  • Check for go.mod (Go)
  • Check for Gemfile (Ruby), Cargo.toml (Rust), composer.json (PHP)

Step 2: Check Tool Availability

  • For each detected ecosystem, verify audit tool is available
  • If tool missing: log warning, skip ecosystem (do not fail)

Phase 2: Audit Execution

Step 1: Run Ecosystem Audits

  • Execute audit command for each detected ecosystem
  • Prefer JSON output for parsing (see references/audit_commands.md)
  • Run audits in parallel where possible

Step 2: Parse Results

  • Normalize findings to common format: package, version, vulnerability ID, severity
  • Extract CVSS score if available

Phase 3: Report Generation

Step 1: Severity Classification

  • Map CVSS scores to severity per references/severity_mapping.md
  • Critical: CVSS 9.0-10.0
  • High: CVSS 7.0-8.9
  • Medium: CVSS 4.0-6.9
  • Low: CVSS 0.1-3.9

Step 2: Group and Sort

  • Group by ecosystem
  • Sort by severity (Critical first)
  • Include vulnerability count summary

Step 3: Build Report

  • Include package name, current version, fixed version
  • Include vulnerability ID (CVE/GHSA/OSV)
  • Do NOT include exploit details

Phase 4: Fix Recommendations

Step 1: Classify Fix Type

  • Patch update (safe auto-fix)
  • Minor update (usually safe)
  • Major update (manual review required)
  • No fix available (document and monitor)

Step 2: Generate Recommendations

  • For each vulnerability: suggest fix command
  • Flag breaking changes if major version bump
  • Note if fix requires code changes

Step 3: Return Results

  • Return structured report to orchestrator
  • Include summary: packages audited, vulnerabilities found, by severity

Critical Rules

  1. Never auto-fix major versions - may introduce breaking changes
  2. Verify lock file integrity - regenerate if corrupted
  3. Respect severity thresholds - per environment (see references/severity_mapping.md)
  4. Document unfixable vulns - add to known issues with review date
  5. No exploit code - report IDs only, not exploitation details

Definition of Done

  • All detected ecosystems audited
  • Findings classified by severity with CVSS mapping
  • Fix recommendations provided (safe vs manual)
  • Report in normalized format returned
  • Critical vulnerabilities prominently flagged
  • Lock file integrity verified

Reference Files

FilePurpose
references/audit_commands.mdEcosystem-specific audit commands
references/severity_mapping.mdCVSS to severity level mapping
references/ci_integration_guide.mdCI/CD integration guidance

Version: 2.0.0 Last Updated: 2026-01-10

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

29.12%
按下载量换算28

Gemini CLI

21.6%
按下载量换算21

Codex

18.11%
按下载量换算18

OpenCode

12.05%
按下载量换算12

Antigravity

6.88%
按下载量换算7

windsurf

3.11%
按下载量换算3

安全审计

暂无安全审计结果可展示。

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。

来源信息

继续浏览同类 Skills