Skip to content

v1.0.0

Latest
Latest

Choose a tag to compare

View all tags
@fb0sh fb0sh released this 02 Jun 03:32
· 33 commits to main since this release
v1.0.0
cfe1d56

Pentester v1.0.0

基于 AI Agent 的 PTES 渗透测试自动化框架

概述

Pentester 是一个由 AI Agent 驱动的渗透测试自动化框架，严格遵循 PTES（Penetration Testing Execution Standard） 标准，覆盖从 Pre-engagement 到 Reporting 的全生命周期。Agent 自动拉起 Kali Linux 容器，按 PTES 七阶段规范逐步执行，最终生成结构化渗透测试报告。

核心功能

  • PTES 七阶段全流程 — Pre-engagement → Intelligence Gathering → Threat Modeling → Vulnerability Analysis → Exploitation → Post Exploitation → Reporting，每阶段含 L1/L2/L3 三级详细流程
  • AI Agent 工作流 — 交互式参数采集（一问一答、编号选项）、合规验证、同阶段多 sub-agent 并行派发
  • 知识库 — .pentester/knowledge/ 下的结构化、可检索、可复用知识库，285+ 条索引，PTES 阶段关联导航与标签系统
  • GitHub 参考资料 — 101 MB 精选 P0/P1 安全研究仓库原文，离线可用
  • 精炼聚类 — 29 个专题桶（24 MB）+ 174K 深度去重聚类（77 MB），涵盖 PoC/利用/工具参考
  • 字典集合 — PentesterSpecialDict 专项字典集：SQL 注入、XSS、SSRF、SSTI、目录枚举、密码爆破、子域名等
  • MCP 工具集成 — 注册表式 MCP 工具发现与阶段内自动调用
  • 技能系统 — 模块化技能：gen-report、knowledge-base、init-target、enum-services、update-solution、mcp-manager
  • 容器隔离 — 全部攻击操作在 Docker 容器内执行，不暴露宿主机
  • 报告生成 — 自动输出渗透测试报告，附带 6 维度 × 24 项质量检查清单

项目结构

pentester/
├── .pentester/                  # 框架核心
│   ├── ROLE.md                  # Agent 角色与行为规则
│   ├── PTES.md                  # 7 阶段路由表
│   ├── sections/                # 各阶段流程（L1/L2/L3）
│   ├── skills/                  # 可调用技能
│   ├── knowledge/               # 可复用知识库
│   ├── wordlists/               # Fuzz/爆破字典
│   ├── mcp/                     # MCP 工具注册表
│   ├── solutions.md             # 累积解决方案库
│   └── REPORT-TEMPLATE.md       # 报告模板
└── target/                      # 单目标测试数据

快速开始

pi
> 对 192.168.39.3 执行渗透测试

Agent 会自动交互确认参数、拉起容器、按 PTES 阶段推进、最后生成报告。

文档

English

Pentester v1.0.0

AI Agent-driven penetration testing framework following the PTES standard

Overview

Pentester is an automated penetration testing framework powered by an AI Agent. It strictly follows the PTES (Penetration Testing Execution Standard) covering the full lifecycle from pre-engagement to reporting. The Agent automatically launches Kali Linux containers, executes each of the 7 PTES phases step by step, and produces a structured penetration test report.

Key Features

  • 7 PTES Phases — Pre-engagement → Intelligence Gathering → Threat Modeling → Vulnerability Analysis → Exploitation → Post Exploitation → Reporting, each with detailed L1/L2/L3 procedures
  • AI Agent Workflow — Interactive parameter collection (one question at a time, numbered options), compliance verification, parallel sub-agent dispatch within the same phase
  • Knowledge Base — Structured, searchable, reusable knowledge under .pentester/knowledge/ with 285+ indexed entries, PTES-linked navigation, and tag system
  • GitHub References — 101 MB of curated P0/P1 security research repositories, fully offline-ready
  • Refined Corpus — 29 topic buckets (24 MB) and 174K de-duplicated clusters (77 MB) for PoC/exploit/tool reference
  • Wordlist Collection — PentesterSpecialDict with specialized dictionaries for SQLi, XSS, SSRF, SSTI, directory enumeration, password attacks, DNS subdomain, and more
  • MCP Tool Integration — Registry-based MCP tool discovery and automatic invocation per phase
  • Skill System — Modular invocable skills: gen-report, knowledge-base, init-target, enum-services, update-solution, mcp-manager
  • Container Isolation — All offensive operations execute inside Docker containers; zero host machine exposure
  • Report Generation — Auto-generated penetration test report with 6-dimension × 24-item quality checklist

Project Structure

pentester/
├── .pentester/                  # Framework core
│   ├── ROLE.md                  # Agent identity & behavior rules
│   ├── PTES.md                  # 7-phase routing table
│   ├── sections/                # Phase procedures (L1/L2/L3)
│   ├── skills/                  # Invocable agent skills
│   ├── knowledge/               # Reusable knowledge base
│   ├── wordlists/               # Fuzz/brute-force dictionaries
│   ├── mcp/                     # MCP tool registry
│   ├── solutions.md             # Cumulative solutions library
│   └── REPORT-TEMPLATE.md       # Report template
└── target/                      # Per-target findings

Quick Start

pi
> Run penetration test against 192.168.39.3

The Agent handles the rest — interactively confirms parameters, launches the container, progresses through PTES phases, and generates the final report.

Assets 3
Loading