Security & AI safety

Pentesting & offensive security

763 agents · most starred first

Pentesting & offensive security, page 3 of 8

legion★ 128

Scrappy assistant that automates web3 bug hunting workflows. Tracks ongoing bug bounties and launches autonomous agents that can perform arbitrary tasks.

pentestkit★ 127

Autonomous multi-agent pentest framework — plans, exploits, verifies (proof-required), CVSS-scores and writes client-ready reports. 104/104 (100%) on the XBOW validation benchmarks, powered by Kimi K3.

aimy-skill★ 120

aimy-skill 是一个轻量级、可嵌入 AI Agent 的渗透测试辅助技能,专为授权环境下的自动化信息收集和基础漏洞探测设计。它可以被 AI 助手(如 AutoGPT、LangChain 应用)调用,也可以独立命令行运行。

cascade-protocol-dissector★ 114

Windsurf Cascade Bypass Methods 2026: Advanced Protobuf & MITM Exploit Repo

rea★ 113

The reverse-engineering framework you can run anywhere, with nothing installed, and actually trust.

AIRTBench-Code★ 111

Code Repository for: AIRTBench: Measuring Autonomous AI Red Teaming Capabilities in Language Models

RiftX★ 110

Pentest anything you want

sqlmap-skynet★ 109

SQLMap with Autonomous AI, phased workflows, RAG memory, and MCP Agent Tools

ai-security-arsenal★ 109

A collection of skills, agents, commands, and workflows for security researchers. Compatible with Claude Code, Claude Desktop, OpenCode, and other AI coding tools.

Whomrx-gpt★ 108

This is a Whomrx GPT which is related to the worm GPT and the natural enemy of the chat GPT {UPDATE 2025}

cybersentry★ 108

🤖 Autonomous AI-powered ethical hacking agent powered by Llama 3.1 70B on NVIDIA NIM

ares★ 106

Ares is an autonomous security operations platform where LLM-driven red and blue team agents operate against each other on live infrastructure, enabling realistic evaluation of attack and defense.

dsh-pentester★ 104

基于 DeepSeek Harness 的多 Agent PTES 渗透测试编排插件,支持自动化侦察、漏洞分析、验证与报告,使用 Docker/Kali 隔离工具箱 | Multi-agent PTES penetration testing plugin for DeepSeek Harness with automated recon, vulnerability analysis, validation, reporting, and Docker/Kali toolbox

open-sesame★ 104

🔓 An AI web-pentest skill for Claude Code. Point it at a site you're authorized to test — it maps the attack surface, works through the OWASP WSTG, validates findings, and writes dev + stakeholder reports (PDF).

Hx0-HawkEye★ 101

面向 Chrome/Firefox 的轻量级浏览器安全工作台:集成抓包、拦截改包、重放/Fuzz、编解码工作台、敏感信息检测、HawkEye MCP 与浏览器级 Agent。(A lightweight Chrome/Firefox browser security workbench for capture, interception, replay/Fuzz, codec toolkit, sensitive-data detection, HawkEye MCP, and browser-level Agent workflows.)

Cyber-Zero★ 100

Cyber-Zero: Training Cybersecurity Agents Without Runtime

exploitation-validator★ 100

A prompt-based pipeline for finding, validating, and proving vulnerabilities using LLM sub-agents.

Threatswarm★ 100

27 scope-enforced AI agents that run the full pentest kill-chain (recon → exploit → post-ex → DFIR → report) as a one-command Claude Code plugin. Backed by 754 MITRE-mapped skills.

cybersec-toolkit★ 100

One command installs 670+ security tools on Debian/Ubuntu/Kali, Fedora, Arch, openSUSE and Termux. Its authorization-gated MCP server runs them in a disposable Kata Containers VM sandbox for Claude Code, Codex, Gemini CLI, OpenCode and other MCP clients. 872 agent skills for CTF, pentesting, bug bounty, DFIR and red/blue teams.

DecodingTrust-Agent-Platform★ 100
kernagent★ 99

Headless AI agent for deterministic reverse engineering.

awesome-ai-reverse-engineering★ 99

AI 辅助逆向工程工具清单,收录 MCP、Skill、插件、Agent 与工作站,覆盖 Web、二进制、移动端、浏览器自动化和流量分析。

auto-pen-bench★ 98

This repo contains the codes of the penetration test benchmark for Generative Agents presented in the paper "AutoPenBench: Benchmarking Generative Agents for Penetration Testing". It contains also the instructions to install, develop and test new vulnerable containers to include in the benchmark.

DecodingTrust-Agent★ 98
Dark-Agent★ 98

Mythic C2 agent with a full in-memory BOF loader for macOS and Linux

Maverick★ 97

Adaptix C2 agent using Crystal Palace PIC linker and PICO module system

hacker-bob★ 97

Claude Code bug bounty agent. Install in current workspace: npx -y hacker-bob-cc@latest install "$PWD" then restart Claude and run /bob-hunt target.com

bughunter-ai★ 97

Autonomous Bug Bounty Hunting Framework powered by Claude Code. 20 AI agents, state-machine orchestration, Burp Suite MCP, credential vault, LLM security track. Type 'hunt target.com' and let AI find the bugs.

hunxiang★ 97

浑象 AI agent CTF 靶场竞赛平台

ARTEX★ 97

ARTEX :AI 自主渗透测试系统(Go + Next.js),基于 Autumn-27/ARTEX v0.3.15(e6ec569,AGPL-3.0)。原版 v0.3.15 下载说明与恢复指引见 README。

XuanMu-RedTeam-Agent★ 96

玄幕安全团队开源汉化和修改的无需docker可在Kali Linux上直接运行的红队AI agent。

ZhiShi★ 96

安全研究领域的 agent harness

rea-engenharia-reversa★ 96

REA (Reverse Engineer Anything) explicado em PT/EN/ES: engenharia reversa com agentes de IA + Ghidra — links, casos, avaliação e cuidados

gpg_reaper★ 95

GPG Reaper - Obtain/Steal/Restore GPG Private Keys from gpg-agent cache/memory

HavocNotion★ 94

A simple ExternalC2 POC for Havoc C2. Communicates over Notion using a custom python agent, handler and extc2 channel. Not operationally safe or stable, built as a PoC to showcase Havoc C2's modular C2 channel interface.

github-C2★ 94

网页控制台 + Rust Agent。Cloudflare Worker 主通道,GitHub Issue 备用。MCP接入AI,全自动渗透必备。

hackmyagent★ 93

Metasploit for AI agents: scan, attack, and fix AI agents and MCP servers. Open source security toolkit.

antares★ 93

Antares: a self-hosted AI pentest agent and assistant in one binary. Authorized penetration testing with security specialists, 7,000+ procedures and reports, plus files, shell, a real browser, memory and RAG.

Starburst★ 93

PIC shellcode agent based on Stardust framework and Crystal Palace

awesome-ai-reverse★ 91

AI 辅助逆向工程工具清单,收录 MCP、Skill、插件、Agent 与工作站,覆盖 Web、二进制、移动端、浏览器自动化和流量分析。

opencode-pentester★ 91

AI-powered red teaming, and pentesting for web applications via opencode : defensive code audit (17 vulnerability categories) + offensive penetration testing (63 attack categories, 15 agents, 11 domains).

genxlm★ 90

A simple script to generate JScript code for calling Win32 API functions using XLM/Excel 4.0 macros via Excel.Application "ExecuteExcel4Macro"

pwn★ 90

PWN is an open security automation framework that aims to stand on the shoulders of security giants, promoting trust and innovation.

ares★ 90

AI Robustness Evaluation System

Kali_Hack_Agent★ 89

一个基于三层 Agent 架构的容器代码执行系统,专门用于在 Kali Linux Docker 容器中安全、自动地执行渗透测试工具。附赠配置好的hexstrike-ai docker

MM-Plan★ 89

[ICLRW 2026 Best Short Paper Award] Visual Exclusivity Attacks: Automatic Multimodal Red Teaming via Agentic Planning

ghidra-ai-bridge★ 87

AI-powered Ghidra query interface for reverse engineering agents

PyHmmm★ 86

Simple PoC Python agent to showcase Havoc C2's custom agent interface. Not operationally safe or stable. Released with accompanying blog post as a tutorial sample

utf8-overlong-agent★ 86

使用 agent 实现反序列化 utf8 overlong

IAT_API★ 86

Assembly block for finding and calling the windows API functions inside import address table(IAT) of the running PE file.

AI-System-Prompts★ 86

XBot - Advanced AI Cybersecurity Agent | Gemini system prompt for automated penetration testing and security assessments

aster★ 86

基于 ReAct 框架的安全分析 Agent,在终端中完成代码审计、渗透测试、主机防护

OpenSecurity★ 86

AI 驱动的多领域安全分析 Agent 平台 —— 让 LLM 端到端完成安全分析

BoxPwnr-Traces★ 85

LLM agent solving traces, leaderboards, and benchmark results across security CTF and hacking platforms

BusyBOF★ 85

Busybox-style Beacon Object Files for *nix post-exploitation. Reimplements common Unix utilities as BOFs for use in stripped environments (Docker containers, Kubernetes pods, minimal VMs) where no binaries exist beyond the agent / implant.

pwner★ 85

Autonomous agentic network penetration-testing tool (authorized use only)

pypykatz_agent_dn★ 84

Pypykatz agent implemented in .NET

wormxgpt★ 84

No limits. No filters. No restrictions. WormXGPT is a unified AI tooling suite containing both a premium Hacker-themed React Web Dashboard and an advanced Unfiltered CLI agent. It features 150+ tools, multi-server MCP integration, auto-fallback across 30+ providers, and local workspace integration.

LyScript★ 84

An AI driven debugging assistant specially designed for x32/x64dbg, it helps to double the efficiency of reverse engineering, vulnerability mining, and malware analysis, making it your intelligent copilot.

RedTeam-Agent★ 83

RedTeam-Agent: AI-Powered Autonomous Red Team Framework via Model Context Protocol. AI红队与内网渗透自动化框架,支持 gogo, fscan, httpx, nuclei, impacket, playwright 等 15+ 渗透工具,让 LLM 直接化身安全审计黑客。

awesome-security-agent-harnesses★ 83

AI pentesting tools, AI SOC and autonomous SOC agents, purple team and red vs blue harnesses, and security agent harnesses for code audit, vulnerability discovery, fuzzing, and reverse engineering. Curated open-source code, cyber ranges and test environments, MCP servers, and benchmarks.

agent-python★ 80

Official python agent for using the distributed hashcracker Hashtopolis

exploitgym★ 79

ExploitGym is a large-scale, realistic benchmark built from real-world vulnerabilities designed to evaluate AI agents' ability to develop exploits.

autonomous-offensive-llm-handbook★ 78

Step-by-step handbook for building autonomous security agents with clear limits, runnable lessons, diagrams, and honest evidence.

secbot★ 77

AI-powered authorized security testing workspace. Maintained on the release branch (NestJS + Ink TUI + Web/Desktop).

bug-reaper★ 76

Web2 bug bounty Agent Skill — evidence-based, no AI slop. Covers 18 vulnerability classes across HackerOne, Bugcrowd, Intigriti, and YesWeHack.

supabase-pentest-skills★ 76

24 AI Agent Skills for professional security auditing of Supabase applications. Detection, key extraction, RLS testing, storage audit, IDOR detection, and comprehensive reporting. Works with Claude Code, Cursor, Windsurf, and 30+ AI agents.

cheatengine-mcp-tcp-bridge★ 76

TCP-enhanced fork of Cheat Engine MCP Bridge. Connect Cursor, Copilot and Claude AI directly to local or remote Cheat Engine via TCP. Automate reverse engineering, pointer scanning, and memory analysis using natural language.

autopentest★ 75

CLI enabling more autonomous black-box penetration tests using Large Language Models (LLMs)

0sec★ 73

[RESEARCH PREVIEW] The open & extensible cybersecurity harness - by the Swiss Applied AI & Cybersecurity Research Lab.

tchkiller★ 72

基于 claude-agent-sdk 的智能渗透测试 Agent,用于参与第二届 TCH 智能渗透挑战赛。

NetSecGame★ 71

An environment simulation for networks security tasks for development and testing AI based agents. Part of AI Dojo project

argus-validation-benchmarks★ 70

Self-contained, Dockerized offensive security challenges for evaluating AI-powered penetration testing agents. Covers modern tech stacks (Node.js, Python, Go, Java, PHP, Ruby) across diverse vulnerability classes and target environments from basic injection to multi-step exploit chains, sandbox escapes, and defense-enabled environments.

PHANTOM★ 70

👻 AI-Powered Pentesting Command Center — Autonomous security testing with real-time streaming, self-improving AI, unlimited tool iterations,

redteam-ai-benchmark★ 67

Red Team AI Benchmark: Evaluating Uncensored LLMs for Offensive Security

claude-active-directory★ 66

AI agent harness for Active Directory offensive security — 8 skill domains, 13 slash commands, 7 agents, ROE-safe orchestration. Part of DoOS by Evaluris Solutions Labs.

templates-extender★ 65

Templates for developing your own listeners and agents for AdaptixC2.

PHANTOM★ 64

👻 AI-Powered Pentesting Command Center — Autonomous security testing with real-time streaming, self-improving AI, unlimited tool iterations,

Vigilo★ 64

An AI hacker for Web3 Smart Contract. for bug bounties, Audit contest, offensive security research, and real-world exploit thinking.

argus-validation-benchmarks★ 64

Self-contained, Dockerized offensive security challenges for evaluating AI-powered penetration testing agents. Covers modern tech stacks (Node.js, Python, Go, Java, PHP, Ruby) across diverse vulnerability classes and target environments from basic injection to multi-step exploit chains, sandbox escapes, and defense-enabled environments.

Cyber-Frost-Harness★ 64

Native red, blue, and purple security-agent harness for CYBER-FROST

claude-mythos★ 63

Prompt framework that makes LLMs discover real, exploitable vulnerabilities. Six specialized agents covering recon, hunting, adversarial chaining, exploit validation, triage, and AI security.

awesome-ai-pentest★ 63

Curated list of open-source AI-assisted penetration testing tools, frameworks, CTF agents, and benchmarks

WormGPT★ 62

WormGPT is an experimental AI project designed strictly for educational, cybersecurity research, and defensive testing purposes only. It is intended to help developers, researchers, and students understand AI behavior, prompt injection risks, and security vulnerabilities in language models.

find-cve-agent★ 62

CVE hunting harness for Claude Code - 20 skills, 5-agent team, systematic vulnerability research with false positive elimination

dhunter★ 62

AI 驱动的自主渗透测试平台:输入目标,AI agent 自动完成侦察→规划→主动测试→漏洞验证→报告生成。黑板引擎+多 worker+SRC 验收门禁。仅供学术与安全研究使用。

PenGym★ 61

PenGym: Pentesting Training Framework for Reinforcement Learning Agents

EvoSynth★ 61

EvoSynth is a SOTA automated LLM red-teaming framework that evolves executable, code-level jailbreak attack methods using multi-agent synthesis, evolutionary search, and self-correction.

agentic-ida-pro★ 61

LLM & IDA 主导的逆向分析平台

DVBLab★ 60

This course uses a deliberately vulnerable banking application to demonstrate common security vulnerabilities, their impact, and how to fix them. The application is built with Flask (backend) and React (frontend).

NThread★ 59

Stealthy x64 thread manipulation library for calling functions inside target processes without creating remote threads or installing hooks.

BugHunter-AI★ 59

Cyberpunk AI Pentest Agent — Automated penetration testing framework with AI-assisted analysis, resource-aware task scheduling, and a futuristic GUI.

Online_Tools-AI★ 59

一个面向 Web 版桌面 安全测试的 AI Agent 平台,集成侦察、分析、规划、工具调用与反馈学习,实现自主完成渗透测试工作流An AI Agent platform for web-based security testing, integrating reconnaissance, analysis, planning, tool orchestration, and feedback-driven learning to autonomously execute penetration testing workflows.

phantom-ethical-redteam★ 58

An autonomous agent that acts as a DEF CON-level Certified Ethical Hacker, using tools such as Nuclei, sqlmap, ffuf, Burp, ZAP, and the Social Engineering Toolkit (SET), (labs, CTFs, HackTheBox, TryHackMe, and bug bounties with a written scope).

mobile-security-skills★ 56

A collection of Claude Code and OpenAI Codex Agent Skills for mobile application security testing

AgentFlipper★ 55

An AI agent that controls your Flipper Zero that uses PyFlipper, LightLLM, RAG, and can connect to your local Ollama instance!

mythic_ornn★ 55

LLM-driven generator for Mythic Agents, Payload-Type and C2 Profiles.

FruityC2-Client★ 54

FruityC2 is a post-exploitation (and open source) framework based on the deployment of agents on compromised machines. Agents are managed from a web interface under the control of an operator.

PULSE-C2★ 54

HTTPS C&C Framework with encrypted beacons, web dashboard, and Windows agent.

SCOPE★ 54

AI agent set for cloud security purple teaming, runs inside Claude Code, Gemini CLI, and Codex.

More in Security & AI safety