agentleFS
Bin-H-17Bin-H-17/aegis-security-skill/skills/security-audit-common/SKILL.md

security-audit-common

Skill0 starsChanged 59 days agoMITView on GitHub

本地离线安全审查 Skill 集合 L2 公共层(共享依赖,不直接由用户触发)。承载:编排脚本 (audit_orchestrator / cli_wrappers / report_synthesizer / minimal_fix_engine / ai_copyright_checker / store_compliance / oss_compliance)、报告模板、条款库 references/、 基线 baselines/。被 L0 与各场景子 skill 调用。命令行调用不改上游、本地离线、0 数据出境。

The file

Files
SKILL.md

security-audit-common(L2 公共层)

本 skill 为共享能力层,不单独对用户暴露;被 project-security-audit 与四个场景子 skill 依赖。

目录结构

security-audit-common/
├── SKILL.md
├── scripts/
│   ├── common.py            # 路径/配置/日志/数据模型/CLI 封装
│   ├── cli_wrappers.py      # gitleaks/semgrep/bandit/osv-scanner/pip-audit/reuse 封装与解析
│   ├── report_synthesizer.py# P0–P3 统一报告合成(SECURITY_REPORT.md + findings json)
│   ├── minimal_fix_engine.py# 最小改动修复(本地 Ollama,缺失回退规则)
│   ├── ai_copyright_checker.py # X1–X3 逐平台一票否决
│   ├── store_compliance.py  # 上架清单(play_policy.yaml)
│   ├── oss_compliance.py    # 开源发布 gate
│   └── audit_orchestrator.py# L0 主链路编排入口
├── templates/
│   └── SECURITY_REPORT.md   # 报告规范模板
├── references/
│   ├── platform_terms.yaml  # D6 平台条款快照(X1–X3 等)
│   └── play_policy.yaml     # D5 上架政策快照
└── baselines/
    ├── gitleaks.toml        # gitleaks 基线(extend default + 允许列表)
    └── semgrep_rules/owasp.yml # semgrep OWASP 规则包

环境与路径(可覆盖)

环境变量默认说明
SECURITY_AUDIT_HOME本目录(security-audit-common)脚本/references/baselines/templates 根
SECURITY_TOOLS_HOME~/security-tools6 个 CLI 安装目录(可用环境变量覆盖)
SECURITY_AUDIT_CWD无子进程工作目录(可选)
OLLAMA_HOSThttp://localhost:11434本地 Ollama 推理地址
OLLAMA_MODELqwen2.5-coder:7b最小修复引擎模型

许可合规底线(铁律)

  • 仅命令行调用开源工具,不改上游;借鉴 DeepAudit 工作流但严禁复制其 AGPL-3.0 代码。
  • 所有产物本地落盘,0 数据出境。
  • 工具缺失 → 单工具降级(BD-01),不整体失败。

安装

见 install/install_tools.ps1(Windows)与 install/install_tools.sh(跨平台),将 6 个 CLI 装入 SECURITY_TOOLS_HOME。

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.