Secskills logo

Secskills

Community
Arenbai
secskills

渗透测试实战技能 v1.3.0。覆盖信息收集、全类漏洞发现(注入全家桶/SSRF/文件类/反序列化/SSTI/越权逻辑/CSRF)、漏洞利用、后渗透、免杀全流程。 已剔除无危害噪音项(明文传输/CORS/响应头缺失等),只报告有实际危害证据的漏洞。 当用户给出具体目标 (IP/域名/URL) 且意图是攻击/利用/拿权限时触发。 不触发: 概念讨论、蓝队防御、代码审计、CVE文档查询。

Overview

PublisherArenbai
RepositorySecSkills
Skill namesecskills
Stars
259
Forks
29
Bundled files
40
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 40 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by Arenbai on GitHub. Read the source before you install it.

Installation

Install the Secskills AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/Arenbai/SecSkills/tree/v1.3.0
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/Arenbai/SecSkills.git \
  .claude/skills/secskills
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Secskills in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Secskills on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Secskills is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

渗透测试实战技能

架构: SKILL.md (本文件) → references/ (按需加载) | 覆盖: 信息收集 → Web漏洞 → 主机 → 后渗透 → 免杀

触发条件

触发(全部满足): ① 意图是攻击/利用/拿权限 ② 给出具体目标 (IP/域名/URL) ③ 涉及漏洞测试/提权/横向/免杀

不触发(任一命中): 概念问答("什么是XSS")| 蓝队/应急/日志分析 | 业务bug修复 | 查CVE(→WebSearch)| AI Prompt注入(→secknowledge-skill)| 白盒审计(→code-audit-skill)

行为: 用户给出具体目标后 → 先弹 AskUserQuestion 选择面板(见「测试确认」),再进入流程。

行为准则(全程有效)

  1. ❗ 授权优先 — 利用步骤输出前确认: 授权渗透 | 本人环境。无授权 → 只出分析,不出武器化Payload
  2. ❗ 引用强制 — CVE/Payload 必须引用 references/ 章节或 WebSearch 验证。未覆盖 → ⚠️ UNABLE TO CITE
  3. ❗ 风险标注 — 🔴致命/🔴高危/🟡中危/🟢低危 + 利用条件
  4. ❗ 链式思维 — 优先输出利用链 (A→B→C),非孤立漏洞
  5. ❗ 命令可执行 — 完整可复制,IP/端口用 <target> 占位
  6. ❗ 宁可漏报不可误报 — 每条漏洞必须过下方四道关卡,任一不过 → 直接丢弃
  7. ❗ 禁用破坏性操作 — 验证一律用无害语句(SELECT/延时/版本探测/写测试文件),严禁修改、删除数据或删表等写操作;危害证明靠"读到数据",不靠"改掉数据"

⚔️ 漏洞过滤体系(输出前逐条过审,任一关卡未过 → 丢弃)

第零关:快速预筛(任意命中 → 丢弃)

判定典型场景
⓪ 扫描器原始告警Burp/AWVS/Nessus 自动报告,未手动验证
① 凭经验推测"可能存在SSRF" — 未实际发请求
② 信息收集中间产物端口开放、版本号、子域名列表 — 是素材不是漏洞
③ 安全加固建议"建议开CSP"、"建议升级TLS" — 不算漏洞
④ 纯静态分析读代码推断,从未请求验证

第一关:自检门(逐项 YES/NO,任一 NO → 丢弃)

[ ] ① 实际发出了验证请求?(命令: ___)
[ ] ② 拿到了真实响应数据?(响应关键内容: ___)
[ ] ③ 造成了实际危害?(读了什么/执行了什么/越权了什么: ___)
    ⚠️ "触发报错"≠"造成危害";"能访问"≠"拿到敏感数据"。③为NO无条件丢弃
[ ] ④ 利用链完整可复现?(入口→___→危害,任一步为推测→NO)
[ ] ⑤ 对照黑名单通过?  [ ] ⑥ 满足对应等级最低准入?

第二关:黑名单(命中即丢弃)

类别永远不报为漏洞
响应头缺 CSP/HSTS/X-Frame-Options 等;Cookie 缺 HttpOnly/Secure/SameSite
信息泄露版本号 Banner;JS/HTML 中的路由/注释;无敏感文件的目录列表;robots.txt;phpinfo(无凭据);错误堆栈(不含密码/Token/密钥)
认证会话用户名枚举(响应/时间差异);密码策略不严格(无爆破成功);autocomplete 开启;默认凭据尝试失败;无验证码(无爆破成功);401/403 响应
SSL/网络明文传输/HTTP未跳转HTTPS;自签名证书;弱加密套件/TLS版本低;OPTIONS/TRACE 开启;后台存在但进不去
版本问题组件版本过旧/EOL 但无对应可验证漏洞(版本老 ≠ 有漏洞,必须落到具体 PoC 并实际打出危害)
未验证利用SQL报错但提取不到数据;上传成功但服务器不解析;无回显反射XSS;SSRF内网可达但没拿到数据;任意文件读只读到公开文件
其他纯功能bug;短信轰炸(无资费损失证明);同系统同类型>3个(降级合并);API无频率限制;利用条件极苛刻(物理接触/猜64位随机数)

CORS/CRLF/Host头/缓存投毒/请求走私/GraphQL内省/开放重定向 → 见 references/web-low-value.md,默认不报,除非满足该文件的报告门槛(完整利用链+实际危害)。

第三关:严重等级准入(不满足 → 降级重审,再降到底则丢弃)

等级最低准入(满足至少一项)
🔴致命获取系统权限(RCE/WebShell)|核心DB拖库(≥3类敏感字段)|核心认证绕过直进后台
🔴高危任意密码重置/登录|重要系统SQL注入(有回显)|SSRF拿到内网凭证/云元数据|任意文件读到配置/密钥|越权增删改查敏感信息|本地提权完整链
🟡中危存储XSS(能窃Cookie)|敏感操作CSRF|非核心SQL注入(有回显)|任意文件操作有实际影响|普通越权|弱口令(有登录成功证据)|子域名接管成功
🟢低危反射XSS(有弹窗证据)|有限越权|需特殊条件的信息泄露(SVN等)

高频误判速查(校准用)

你看到的实际判断
"用户名不存在"vs"密码错误"响应不同不是漏洞,UX设计
/admin 返回302跳登录已正确保护,不算未授权
上传.php返回200服务器必须实际解析执行才算
内网IP:端口有HTTP响应必须通过SSRF拿到敏感数据/云凭证才算
4位数字验证码可识别必须实际爆破成功才算

幻觉防护

内容正确做法禁止
CVE编号WebSearch → WebFetch PoC → 实际验证凭记忆编造编号/版本范围,看到版本号直接报CVE
Payload引用 references/ 或搜索验证凭记忆写
无匹配⚠️ UNABLE TO ASSESS: 未覆盖,建议[行动]凭经验断言
黑名单命中跳过包装成"低危"输出

标注: [引用:file:section] · ⚠️ 通用知识 · 💡 方法论推理

🎯 测试确认(用户给出目标时,先弹 AskUserQuestion)

一次性弹出 3 个问题,方向键选择、Enter 确认,完成后立即进入 Step 1:

  1. 授权级别: 🔴 授权渗透(完整利用链)/ 🟡 灰盒(有测试账号)/ 🟢 黑盒(无凭证)/ ⚪ 本人环境(靶场/CTF)
  2. 测试深度: 标准测试(推荐,全量检测+验证)/ 快速扫描 / 深度测试(含横向后渗透)/ 仅信息收集
  3. 测试范围: 仅主目标 / 含子域名 / 含关联资产(C段/同ASN)

有测试账号 → 用户通过"其他"填写。

工作流程

两阶段: [攻击] Step 1+2 → [利用] Step 3+4。利用阶段只能引用攻击阶段已加载的文件,禁止跨阶段新增 reference。

🔍 攻击阶段

Step 1: 信息收集 + 攻击面识别

  • 分类目标 (Web/主机/内网) → 匹配导航索引 → 攻击面: 端口/服务/Web入口/认证/API/子域名
  • 指纹组件+版本 → 标记为「历史漏洞候选」,Step 2a 优先处理
  • ✅ Step1: 类型={X}, 攻击面={N}项, 历史漏洞候选={P}个

Step 2a: 历史漏洞匹配(⭐ 最高投入产出比,优先于通用检测)

  1. 提取指纹组件(有/无版本号都要)→ 并行搜索: "[组件] [版本] CVE exploit" / "[组件] RCE 漏洞 PoC" / site:exploit-db.com [组件] / site:github.com [组件] exploit
  2. 有公开 PoC → WebFetch 加载 → 对目标实际验证;无 PoC → 跳过,不猜测
  3. 无版本号组件(Shiro/WebLogic 等)→ 用特征检测方法直接验证,不需精确版本
  4. 同样过全部自检门 — 版本匹配 ≠ 漏洞存在;验证失败记录原因,不进报告
  5. 命中致命/高危 → 立即停止深入,先确认记录

Step 2b: 通用漏洞检测

  • 第一梯队必测 → 第二梯队验证后报 → 低价值类查 web-low-value.md 后默认跳过
  • 仅输出已过自检门的确认漏洞(等级+前提+检测命令+响应证据),禁止输出"漏洞假设"列表
  • ✅ Step2: 历史漏洞={X}条(验证{Y}/排除{Z}), 通用确认={K}条, 过滤噪音={N}条

⚔️ 利用阶段

Step 3: 武器化 + 利用链 + 报告

  • 从已加载文件取利用 Payload → 优先组合利用链 (A→B→C→RCE),每条引用具体 section
  • 每条漏洞/链必须「能打出危害+有链+有证据」,否则跳过 → 按下方模板生成报告(含修复方案)

Step 4: 后渗透(按需) — 提权/横向/凭据窃取/域渗透/持久化/痕迹清理

报告输出格式(强制)

仅两部分。无确认漏洞时只输出:✅ 测试完成,未发现可利用漏洞。 禁止输出: "可能存在"/"疑似"/"建议检查"/安全配置建议/扫描器原始告警。

### 🔴/🟡/🟢 [漏洞名称] — [目标URL/端点]
**危害**: [一句话,实际造成的危害,非推测]
**证据**: 请求: [完整命令] | 响应: [脱敏关键内容]
**利用链**: [入口] → [步骤] → [危害]
**自检清单**: ①请求YES ②响应YES ③危害YES ④链完整YES ⑤黑名单通过YES ⑥等级准入YES
**修复方案**: [具体可操作,非泛泛而谈]

测试摘要: 目标/时间/确认漏洞N条(分级计数)/已排除噪音M条

场景导航索引

信息收集

场景reference
端口扫描+服务识别references/info-port-scan.md
子域名枚举+接管检测references/info-subdomain.md
目录/文件爆破references/info-dir-brute.md
Web指纹/OSINTreferences/info-fingerprint.md / references/info-osint.md

Web 漏洞 — 检测

🔴第一梯队必测必报 | 🟡第二梯队有实际危害才报 | ⚫低价值类默认不报

梯队场景reference关键检测
🔴SQL 注入references/web-sqli.md闭合/报错/延时/Union/读写文件
🔴NoSQL/LDAP/XPath/EL 注入references/web-injection-ext.md操作符绕过/盲注提取
🔴命令执行 RCEreferences/web-rce.md拼接符/回显/不出网
🔴SSRFreferences/web-ssrf.md内网探测/云元数据/Gopher
🔴文件上传references/web-upload.md后缀/内容/条件竞争
🔴文件包含/路径遍历references/web-lfi-path.md伪协议/日志投毒/截断
🔴XXEreferences/web-xxe.md文件读取/Blind/外带DTD
🔴反序列化references/web-deser.mdPHP/Java/Python gadget
🔴SSTI 模板注入references/web-ssti.mdJinja2/Twig/FreeMarker
🔴越权/逻辑/JWT/CSRFreferences/web-auth-logic.mdIDOR/支付/密码重置/JWT攻击/会话/CSRF
🟡XSSreferences/web-xss.md反射/存储/DOM(报告门槛见自检门)
🟡目录遍历/敏感文件references/web-dir-traversal.md仅读到非公开敏感文件才报
🟡竞争条件references/web-race-condition.md仅成功绕过业务限制才报
⚫CORS/CRLF/Host头/缓存投毒/走私/GraphQL/明文传输/开放重定向references/web-low-value.md默认跳过,门槛见文件

Web 漏洞 — 利用阶段

场景reference
WAF/IDS 绕过(利用阶段通用)references/web-waf-bypass.md

各漏洞利用 Payload 在对应 reference 的「利用」section 中。

主机与后渗透

场景reference
密码爆破references/host-brute.md
Linux 提权references/post-linux-privesc.md
Windows 提权references/post-win-privesc.md
凭据窃取+横向references/post-credentials.md
域渗透references/post-ad.md
免杀: Shellcode混淆+加载器references/evasion-shellcode.md

零结果处理

情况动作
目标不可达❌ UNABLE TO ASSESS: 目标无响应
Reference 未覆盖⚠️ UNABLE TO CITE: 建议 WebSearch [关键词]
无授权仅检测方法,不出武器化链
WAF拦截加载 web-waf-bypass.md
利用失败检查版本→防护→替代Payload

路由边界

诉求路由
渗透/红队/提权本 Skill
AI/LLM 安全测试secknowledge-skill
白盒代码审计code-audit-skill
查CVE/文档WebSearch

v1.3.0 | 26个reference | 变化: 6个低价值文件合并为 web-low-value.md;新增 web-injection-ext.md(NoSQL/LDAP/XPath/EL);子域名接管、JWT算法攻击入库;删除 tools- 工具速查(665行);补 CSRF/CSWSH、宽字节注入;WAF 绕过闭环(sqli/xss → web-waf-bypass.md 分层流程);SKILL.md 428→~230行*

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Secskills AI skill do?

渗透测试实战技能 v1.3.0。覆盖信息收集、全类漏洞发现(注入全家桶/SSRF/文件类/反序列化/SSTI/越权逻辑/CSRF)、漏洞利用、后渗透、免杀全流程。 已剔除无危害噪音项(明文传输/CORS/响应头缺失等),只报告有实际危害证据的漏洞。 当用户给出具体目标 (IP/域名/URL) 且意图是攻击/利用/拿权限时触发。 不触发: 概念讨论、蓝队防御、代码审计、CVE文档查询。

Why use Secskills on TypingMind?

Because you install it once and use it with any model. Secskills is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Secskills in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Arenbai/SecSkills/tree/v1.3.0. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Secskills?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Secskills?

As many as you like. As long as a model supports skills, you can use Secskills with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Secskills AI skill free?

Yes. It is published on GitHub by Arenbai under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇