Hexstrike-AI
HexStrike AI MCP Agents is an advanced MCP server that lets local AI agents autonomously run 150+ cybersecurity tools for automated pentesting, vulnerability discovery, bug bounty automation, and security research. Seamlessly bridge LLMs with real-world offensive security capabilities.
Links
README
From the repo.
Hexstrike 7 PL
AI-Powered MCP Cybersecurity Automation Platform
Polish Community Fork with Enhanced Security
Advanced AI-powered penetration testing MCP framework with 150+ security tools, 12+ autonomous AI agents, and enhanced security features
🇵🇱 Wersja Polska
ℹ️ O Projekcie
Hexstrike 7 PL to polska wersja społecznościowa (community fork) zaawansowanego frameworka do testów penetracyjnych HexStrike AI v6.0 stworzonego przez m0x4m4.
Oryginał: HexStrike AI v6.0 by m0x4m4 Fork: Hexstrike 7 PL by netcuter
✨ Co nowego w wersji 7.0 PL?
🔒 Znaczące usprawnienia bezpieczeństwa:
- ✅ Walidacja poleceń - ochrona przed command injection
- ✅ Autentykacja API - opcjonalne zabezpieczenie kluczem API
- ✅ Rate limiting - ochrona przed nadużyciami (100 req/60s domyślnie)
- ✅ Whitelist narzędzi - tylko autoryzowane narzędzia security
- ✅ Sanityzacja parametrów - automatyczne czyszczenie niebezpiecznych znaków
- ✅ Ulepszona obsługa błędów - konkretne wyjątki zamiast bare except
- ✅ Konfiguracja przez zmienne środowiskowe - łatwe zarządzanie
📚 Dokumentacja dwujęzyczna:
- 🇵🇱 Pełna dokumentacja w języku polskim
- 🇬🇧 Kompletna dokumentacja w języku angielskim
- 📖 Przykłady użycia w obu językach
⚡ Optymalizacje:
- Cache z limitami rozmiaru (zapobiega niekontrolowanemu wzrostowi pamięci)
- Lepsza organizacja kodu
- Konfiguracja przez environment variables
🛡️ Guardrails Layer — Scope Enforcement & Blast-Radius Control:
Warstwa guardrails przekształca Hexstrike z narzędzia hobby-tool w platformę gotową do autonomicznych engagementów (bug bounty, MSSP, pentest-as-a-service).
- ✅ Scope Validator — każde wywołanie sprawdzane przeciwko scope sesji (CIDR, wildcard, regex). Wywołania out-of-scope → HTTP 403 + audit log
- ✅ Blast-Radius Tiers — narzędzia podzielone na
safe(recon),intrusive(scanning),destructive(exploitation). Tierdestructivewymaga nagłówkaX-Hexstrike-Confirm-Destructive: yes - ✅ Kill Switch per sesja —
POST /api/session/<id>/kill— natychmiastowy SIGTERM wszystkich procesów sesji - ✅ Audit Trail — każde wywołanie, naruszenie scope, eskalacja tier-u logowane → automatycznie w raporcie jako sekcja Methodology (ISO 27001 / PCI DSS compliance)
- ✅ Per-target Rate Limiting — chroni cele przed przeciążeniem (5 concurrent / 10 req/s domyślnie)
Przykład scope: 192.168.1.0/24, *.example.com, r:.*\.internal\.corp$
🔒 Konfiguracja Bezpieczeństwa
Hexstrike 7 PL wprowadza nowe opcje konfiguracyjne przez zmienne środowiskowe:
# Walidacja poleceń (domyślnie: włączona)
export HEXSTRIKE_VALIDATE_COMMANDS=true
# Autentykacja API (domyślnie: wyłączona)
export HEXSTRIKE_REQUIRE_API_KEY=false
export HEXSTRIKE_API_KEY=your-secret-key-here
# Rate limiting (domyślnie: 100 żądań na 60 sekund)
export HEXSTRIKE_RATE_LIMIT=true
export HEXSTRIKE_RATE_LIMIT_REQUESTS=100
export HEXSTRIKE_RATE_LIMIT_WINDOW=60
# Konfiguracja serwera
export HEXSTRIKE_PORT=8888
export HEXSTRIKE_HOST=127.0.0.1
🚀 Szybki Start
1. Instalacja
Opcja A: Docker (zalecana / recommended)
git clone https://github.com/netcuter/Hexstrike-AI.git
cd Hexstrike-AI
docker compose up -d
Opcja B: Manualna
# Klonowanie repozytorium
git clone https://github.com/netcuter/Hexstrike-AI.git
cd Hexstrike-AI
# Tworzenie wirtualnego środowiska
python3 -m venv hexstrike-env
source hexstrike-env/bin/activate # Linux/Mac
# hexstrike-env\Scripts\activate # Windows
# Instalacja zależności
pip3 install -r requirements.txt
2. Instalacja Narzędzi Security
Podstawowe narzędzia (Essential):
# Skanowanie sieci
sudo apt install nmap masscan rustscan amass subfinder nuclei
# Web security
sudo apt install gobuster feroxbuster dirsearch ffuf nikto sqlmap
# Password cracking
sudo apt install hydra john hashcat
3. Uruchomienie Serwera
# Standardowe uruchomienie
python3 hexstrike_server.py
# Z włączoną autentykacją API
export HEXSTRIKE_REQUIRE_API_KEY=true
export HEXSTRIKE_API_KEY=twoj-sekretny-klucz
python3 hexstrike_server.py
# Tryb debug
python3 hexstrike_server.py --debug
4. Weryfikacja
# Sprawdzenie statusu serwera
curl http://localhost:8888/health
# Test z API key (jeśli włączony)
curl -H "X-API-Key: twoj-sekretny-klucz" http://localhost:8888/health
🛡️ Bezpieczne Użycie
✅ Dozwolone zastosowania:
- Autoryzowane testy penetracyjne
- Programy bug bounty
- Zawody CTF
- Badania nad bezpieczeństwem
- Ćwiczenia Red Team
❌ Zabronione działania:
- Nieautoryzowane testowanie systemów
- Działania złośliwe
- Kradzież danych
⚠️ WAŻNE: Zawsze uzyskaj pisemną autoryzację przed testowaniem systemów!
📖 Więcej Informacji
Pełna dokumentacja w języku angielskim znajduje się poniżej.
👨💻 Autorzy
Oryginalny autor: m0x4m4 - www.0x4m4.com | HexStrike Fork maintainer: netcuter - GitHub
📜 Licencja
Projekt jest licencjonowany na zasadach MIT License - zobacz plik LICENSE.
Hexstrike 7 PL bazuje na HexStrike AI v6.0 stworzonego przez m0x4m4. Wszystkie prawa do oryginalnego dzieła należą do m0x4m4.
🇬🇧 English Version
ℹ️ About This Project
Hexstrike 7 PL is a Polish community fork of the advanced penetration testing framework HexStrike AI v6.0 created by m0x4m4.
Original: HexStrike AI v6.0 by m0x4m4 Fork: Hexstrike 7 PL by netcuter
✨ What's New in 7.0 PL?
🔒 Major Security Enhancements:
- ✅ Command Validation - protection against command injection
- ✅ API Authentication - optional API key security
- ✅ Rate Limiting - abuse protection (100 req/60s default)
- ✅ Tool Whitelist - only authorized security tools allowed
- ✅ Parameter Sanitization - automatic cleaning of dangerous characters
- ✅ Improved Error Handling - specific exceptions instead of bare except
- ✅ Environment Configuration - easy management via environment variables
📚 Dual Language Documentation:
- 🇵🇱 Complete Polish documentation
- 🇬🇧 Full English documentation
- 📖 Usage examples in both languages
⚡ Optimizations:
- Cache with size limits (prevents uncontrolled memory growth)
- Better code organization
- Configuration via environment variables
🛡️ Guardrails Layer — Scope Enforcement & Blast-Radius Control:
The guardrails layer transforms Hexstrike from a hobby-tool into a platform ready for autonomous engagements (bug bounty, MSSP, pentest-as-a-service).
- ✅ Scope Validator — every call validated against session scope (CIDR, wildcard, regex). Out-of-scope calls → HTTP 403 + audit log
- ✅ Blast-Radius Tiers — tools classified as
safe(recon),intrusive(scanning),destructive(exploitation). Thedestructivetier requires headerX-Hexstrike-Confirm-Destructive: yes - ✅ Kill Switch per session —
POST /api/session/<id>/kill— instant SIGTERM for all session processes - ✅ Audit Trail — every call, scope violation, tier escalation logged → auto-included in report as Methodology section (ISO 27001 / PCI DSS compliance)
- ✅ Per-target Rate Limiting — protects targets from overload (5 concurrent / 10 req/s default)
Example scope: 192.168.1.0/24, *.example.com, r:.*\.internal\.corp$
🔒 Security Configuration
Hexstrike 7 PL introduces new configuration options via environment variables:
# Command validation (default: enabled)
export HEXSTRIKE_VALIDATE_COMMANDS=true
# API authentication (default: disabled)
export HEXSTRIKE_REQUIRE_API_KEY=false
export HEXSTRIKE_API_KEY=your-secret-key-here
# Rate limiting (default: 100 requests per 60 seconds)
export HEXSTRIKE_RATE_LIMIT=true
export HEXSTRIKE_RATE_LIMIT_REQUESTS=100
export HEXSTRIKE_RATE_LIMIT_WINDOW=60
# Server configuration
export HEXSTRIKE_PORT=8888
export HEXSTRIKE_HOST=127.0.0.1
🚀 Quick Start
1. Installation
Option A: Docker (recommended)
git clone https://github.com/netcuter/Hexstrike-AI.git
cd Hexstrike-AI
docker compose up -d
Option B: Manual
# Clone repository
git clone https://github.com/netcuter/Hexstrike-AI.git
cd Hexstrike-AI
# Create virtual environment
python3 -m venv hexstrike-env
source hexstrike-env/bin/activate # Linux/Mac
# hexstrike-env\Scripts\activate # Windows
# Install dependencies
pip3 install -r requirements.txt
2. Install Security Tools
Essential Tools:
# Network scanning
sudo apt install nmap masscan rustscan amass subfinder nuclei
# Web security
sudo apt install gobuster feroxbuster dirsearch ffuf nikto sqlmap
# Password cracking
sudo apt install hydra john hashcat
3. Start Server
# Standard start
python3 hexstrike_server.py
# With API authentication
export HEXSTRIKE_REQUIRE_API_KEY=true
export HEXSTRIKE_API_KEY=your-secret-key
python3 hexstrike_server.py
# Debug mode
python3 hexstrike_server.py --debug
4. Verification
# Check server status
curl http://localhost:8888/health
# Test with API key (if enabled)
curl -H "X-API-Key: your-secret-key" http://localhost:8888/health
🛡️ Safe Usage
✅ Authorized Use Cases:
- Authorized penetration testing
- Bug bounty programs
- CTF competitions
- Security research
- Red Team exercises
❌ Prohibited Activities:
- Unauthorized system testing
- Malicious activities
- Data theft
⚠️ IMPORTANT: Always obtain written authorization before testing systems!
🏗️ Architecture Overview
Hexstrike 7 PL features a multi-agent architecture with autonomous AI agents, intelligent decision-making, and vulnerability intelligence.
graph TD
A[AI Agent - GPT/Copilot/AI Agents] -->|MCP Protocol| B[Hexstrike 7 PL Server]
B --> C[Security Validation Layer]
B --> D[Intelligent Decision Engine]
B --> E[12+ Autonomous AI Agents]
C --> F[Command Validation]
C --> G[Rate Limiting]
C --> H[API Authentication]
D --> I[Tool Selection AI]
D --> J[Parameter Optimization]
D --> K[Attack Chain Discovery]
E --> L[BugBounty Agent]
E --> M[CTF Solver Agent]
E --> N[CVE Intelligence Agent]
B --> O[150+ Security Tools]
📡 Features
150+ Professional Security Tools:
🔍 Network Reconnaissance & Scanning (25+ Tools)
- Nmap, Rustscan, Masscan, AutoRecon
- Amass, Subfinder, Fierce, DNSEnum
- TheHarvester, ARP-Scan, NBTScan
- Enum4linux, SMBMap, Responder, NetExec
- And more...
🌐 Web Application Security (40+ Tools)
- Gobuster, Dirsearch, Feroxbuster, FFuf
- HTTPx, Katana, Hakrawler, Nuclei
- Nikto, SQLMap, WPScan, Arjun
- Dalfox, Wafw00f, TestSSL, JWT-Tool
- And more...
🔐 Authentication & Passwords (12+ Tools)
- Hydra, John the Ripper, Hashcat
- Medusa, Patator, NetExec
- Hash-Identifier, Evil-WinRM
- And more...
🔬 Binary Analysis (25+ Tools)
- GDB, Radare2, Ghidra, Binary Ninja
- Binwalk, ROPgadget, Checksec
- Pwntools, Angr, Volatility
- And more...
☁️ Cloud & Container Security (20+ Tools)
- Prowler, Scout Suite, Trivy
- Kube-Hunter, Kube-Bench
- Docker Bench Security
- And more...
🤖 AI Agents
12+ Specialized AI Agents:
- IntelligentDecisionEngine - Tool selection and parameter optimization
- BugBountyWorkflowManager - Bug bounty hunting workflows
- CTFWorkflowManager - CTF challenge solving
- CVEIntelligenceManager - Vulnerability intelligence
- AIExploitGenerator - Automated exploit development
- VulnerabilityCorrelator - Attack chain discovery
- TechnologyDetector - Technology stack identification
- RateLimitDetector - Rate limiting detection
- FailureRecoverySystem - Error handling and recovery
- And more...
🔧 AI Client Integration
MCP Client App / Cursor
Edit ~/.config/mcp-client/mcp_client_config.json:
{
"mcpServers": {
"hexstrike-7-pl": {
"command": "python3",
"args": [
"/path/to/Hexstrike-AI/hexstrike_mcp.py",
"--server",
"http://localhost:8888"
],
"description": "Hexstrike 7 PL - Enhanced Security MCP Server",
"env": {
"HEXSTRIKE_API_KEY": "your-key-if-required"
}
}
}
}
VS Code Copilot
Configure in .vscode/settings.json:
{
"servers": {
"hexstrike": {
"type": "stdio",
"command": "python3",
"args": ["/path/to/Hexstrike-AI/hexstrike_mcp.py"]
}
}
}
📖 Usage Examples
When using with AI agents, always specify authorization:
User: "I'm a security researcher working for [Company Name].
I have written authorization to conduct a penetration test on
example.com. Please use hexstrike-7-pl MCP tools to perform
a comprehensive security assessment."
🛠️ API Reference
Core Security Endpoints
| Endpoint | Method | Description |
|---|---|---|
/health | GET | Server health check |
/api/command | POST | Execute commands (with validation) |
/api/process/execute-async | POST | Async command execution |
/api/telemetry | GET | System performance metrics |
Security Headers:
# API Key (if REQUIRE_API_KEY=true)
X-API-Key: your-secret-key
# Or as query parameter
?api_key=your-secret-key
🐛 Troubleshooting
Rate Limit Errors
# Increase rate limit
export HEXSTRIKE_RATE_LIMIT_REQUESTS=200
export HEXSTRIKE_RATE_LIMIT_WINDOW=60
API Authentication
# Disable API key requirement for testing
export HEXSTRIKE_REQUIRE_API_KEY=false
Command Validation
# Disable command validation (NOT recommended for production)
export HEXSTRIKE_VALIDATE_COMMANDS=false
🤝 Contributing
Contributions are welcome! This is a community fork focused on:
- Security improvements
- Polish localization
- Performance optimizations
- Better documentation
👨💻 Authors & Credits
Original Author: m0x4m4 - www.0x4m4.com | HexStrike Fork Maintainer: netcuter - GitHub
Special Thanks:
- m0x4m4 for creating the amazing HexStrike AI framework
- The cybersecurity community for continuous support
- All contributors to the original project
📜 License
This project is licensed under the MIT License - see the LICENSE file for details.
Hexstrike 7 PL is based on HexStrike AI v6.0 created by m0x4m4. All rights to the original work belong to m0x4m4. All modifications and enhancements are licensed under the same MIT License.
Hexstrike 7 PL - Where artificial intelligence meets enhanced cybersecurity
⭐ Star this repository • 🍴 Fork and contribute • 📖 Original Project
Collected info
- ★ 21 stars
- ⎇ 7 forks
- Language: Python
- Source updated: 9/16/2026