Skip to content

About

SentinelAI: AI-powered web application security scanner for automated vulnerability assessment and penetration testing.

Topics

Resources

Stars

2 stars

Watchers

0 watching

Forks

Repository files navigation

🛡️ SentinelAI — AI Offensive Security & Vulnerability Arsenal

Python 3.10+ Cross-Platform AppSec AI Engine Copilot MIT License

SentinelAI is an advanced, all-in-one AI-powered offensive security framework, automated vulnerability scanner, and penetration testing arsenal. Combining instant reconnaissance, deep web vulnerability audits, automated threat modeling, CVSS 3.1 calculations, copy-paste developer remediation patches, an interactive AI Security Copilot, and a curated multi-tool launchpad into a seamless terminal and web experience.


 ███████╗███████╗███╗   ██╗████████╗██╗███╗   ██╗███████╗██╗      █████╗ ██╗
 ██╔════╝██╔════╝████╗  ██║╚══██╔══╝██║████╗  ██║██╔════╝██║     ██╔══██╗██║
 ███████╗█████╗  ██╔██╗ ██║   ██║   ██║██╔██╗ ██║█████╗  ██║     ███████║██║
 ╚════██║██╔══╝  ██║╚██╗██║   ██║   ██║██║╚██╗██║██╔══╝  ██║     ██╔══██║██║
 ███████║███████╗██║ ╚████║   ██║   ██║██║ ╚████║███████╗███████╗██║  ██║██║
 ╚══════╝╚══════╝╚═╝  ╚═══╝   ╚═╝   ╚═╝╚═╝  ╚═══╝╚══════╝╚══════╝╚═╝  ╚═╝╚═╝
      ⚡ AI-POWERED OFFENSIVE SECURITY & PENETRATION TESTING ARSENAL ⚡

🚀 Key Highlights & Capabilities

🪟 100% Native Windows & Linux Cross-Platform Architecture

SentinelAI is fully engineered to run seamlessly across Windows 10/11, Linux, and macOS:

  • Windows Package Manager (winget) & Chocolatey (choco): 1-click install industry-standard external tools directly from the terminal console.
  • Smart Path Resolver: Automatically locates tools in C:\Program Files, C:\Program Files (x86), WinGet Links, Chocolatey bin, Scoop shims, and virtual environment Scripts/.
  • Zero-Setup SQLite Mode: Embedded database fallback ensures scans and API servers run natively on Windows without requiring Docker, PostgreSQL, or WSL.
  • Windows Console ANSI & UTF-8: Automatic Virtual Terminal Processing (VT100) initialization guarantees crystal-clear cyberpunk colors, bold borders, and emoji icons in Windows Command Prompt (CMD), PowerShell, and Windows Terminal.
  • Windows Celery Compatibility: Automatically activates the solo pool on Windows to eliminate Linux fork() compatibility errors.

1. 🤖 Next-Gen AI Security Engine & Threat Modeling

  • Multi-Provider LLM Integration: Connects to Groq (llama-3.3-70b-versatile), OpenAI (gpt-4o-mini), Venice AI (llama-3.3-70b), and DeepSeek (deepseek-chat) with automatic multi-provider failover.
  • Offline Heuristic Engine: Built-in comprehensive rule and pattern database guarantees full scoring and remediation capabilities with zero API keys required.
  • Automated Holistic Threat Modeling: Correlates perimeter reconnaissance, open ports, technologies, and vulnerabilities into an executive posture rating (CRITICAL, HIGH, MEDIUM), primary threat vectors, immediate 24-hour actions, and strategic roadmaps.
  • CVSS 3.1 Scoring & Vectors: Calculates standardized CVSS 3.1 vector strings (e.g. CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H) and exploit complexity metrics.
  • Developer Remediation Code Patches: Generates precise, copy-paste defensive patches for Python, PHP, JavaScript, SQL, and Nginx/Apache web server configurations.

2. 💬 Interactive AI Security Copilot (Terminal Chat Mode)

  • Context-Aware Technical Advisory: The Copilot automatically ingests the active target, detected WAF, open ports, and discovered vulnerabilities.
  • On-Demand Penetration Testing Guidance: Ask the Copilot technical questions right inside the terminal:
    • "How do I verify if this SQL injection is vulnerable to time-based exfiltration?"
    • "Generate the secure Nginx configuration to enforce HSTS and CSP."
    • "Explain the business impact of this CORS misconfiguration to an executive."
  • Post-Scan 1-Click Consultation: Launch Copilot directly from the scan summary to immediately interrogate findings.

3. 🧰 Curated Security Tools Arsenal (Inspired by hackingtool)

Instant host environment detection, 1-click auto-installation, live execution streaming, and evidence logging for 26+ industry-standard security tools across 10 categories:

  • Network Discovery & Port Auditing: Nmap, RustScan, Masscan, Netcat (nc)
  • Web Vulnerability & DAST Scanners: Nikto, Nuclei, Wapiti
  • CMS & Framework Auditing: WPScan, CMSmap, Droopescan
  • Directory, File & Parameter Fuzzing: Gobuster, FFuF, Dirsearch, Feroxbuster
  • Database & SQL Injection Auditing: SQLmap, Ghauri
  • OSINT & Subdomain Intelligence: Sublist3r, WhatWeb, Whois, theHarvester
  • DNS & Infrastructure Enumeration: Dnsrecon, Dnsenum
  • SSL / TLS & Cipher Auditing: testssl.sh, SSLyze
  • Secrets & Repository Intelligence: TruffleHog
  • Web Parameter & Route Spidering: ParamSpider

4. ⚡ Multi-Task Automated Workflows (Chained 1-Click Tasks)

  • Perimeter & Infrastructure Recon: DNS + crt.sh Subdomains + 22 Ports + WAF + Security Headers (~10s).
  • Content & Sensitive File Discovery: Web Crawl + 50 Paths + Configs + Backups + LFI (~25s).
  • API & Endpoint Attack Surface: Swagger/OpenAPI + GraphQL + CORS + SSRF Vectors (~20s).
  • Complete Full-Scope Assessment: End-to-end full audit suite with AI correlation and false-positive elimination.

5. 🔍 Standalone Reconnaissance & OSINT Engine

  • Subdomain Discovery: High-speed discovery via Certificate Transparency (crt.sh) & asynchronous DNS resolution.
  • Port Scanner: Multi-threaded TCP connect scanner targeting top operational service ports.
  • WAF & Firewall Fingerprinting: Automatic detection for Cloudflare, AWS WAF, Akamai, Imperva, Sucuri, and ModSecurity.
  • Tech Stack Identification: Identifies Nginx, Apache, WordPress, React, Next.js, Django, Laravel, and exposed server versions.

6. 🎯 Deep Web Application Vulnerability Scanner

  • SQL Injection (SQLi): Error-based, boolean-based, and union-based injection audit.
  • Cross-Site Scripting (XSS): Reflected, DOM-based, and attribute-context script injection checks.
  • Security Misconfigurations: Missing HSTS, CSP, X-Frame-Options clickjacking protection, and CORS reflection.
  • Sensitive File & Information Disclosure: Audits /.env, /.git, backup dumps, Swagger docs, and configuration files.
  • API & Authentication Auditing: Discovers REST/GraphQL endpoints, JWT signing flaws, and IDOR patterns.

7. 📄 Executive & Technical Security Reports

  • Cyberpunk Dark HTML Reports: Executive summary, AI threat model card, risk distributions, CVSS vector badges, and syntax-highlighted code patches.
  • Markdown Reports: Formatted with executive summaries and developer patches, ready for HackerOne, Bugcrowd, or internal wikis.
  • JSON Reports: Machine-readable data for automated DevSecOps pipelines.

⚡ Quick Start

🪟 Windows Quick Start (1-Click Automated Setup)

Option A: 1-Click Batch Installer

Simply double-click install_windows.bat (or run in Command Prompt):

git clone https://github.com/Mazonia/sentinelai.git
cd sentinelai
install_windows.bat

This automatically sets up Python virtual environments, installs requirements, registers global sentinelai commands, and launches the console.

Option B: PowerShell Setup

git clone https://github.com/Mazonia/sentinelai.git
cd sentinelai
.\install_windows.ps1

Once installed, simply type sentinelai or .\sentinelai.ps1 in any terminal!


🐧 Linux & macOS Quick Start

Option A: 1-Click Automated Script

git clone https://github.com/Mazonia/sentinelai.git
cd sentinelai
chmod +x install_linux.sh sentinel.sh
./install_linux.sh

This automatically sets up the Python virtual environment, installs dependencies, registers global sentinelai commands in ~/.local/bin, and launches the framework!

Option B: Manual Setup

git clone https://github.com/Mazonia/sentinelai.git
cd sentinelai
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
pip install -e .
sentinelai

Headless Server Support: On headless Linux VPS/cloud servers (without a desktop GUI/browser), SentinelAI automatically detects the environment and provides direct absolute file paths for HTML/Markdown reports without throwing xdg-open browser launch errors.

2. Interactive Numbered Console (Zero Typing Required)

Simply type sentinelai (or sentinel) — no python prefixes or complex syntax needed:

sentinelai

This launches the guided numbered console:

[1] ⚡ Multi-Task Assessment Workflows (1-Click Chained Presets: Perimeter, Content, API, Full)
[2] 🎯 Full Deep Vulnerability Scan (Crawl, OWASP Audit Suite, AI Threat Model & Report)
[3] 🔍 Reconnaissance & OSINT Engine (DNS, Subdomains, 22-Port Scanner, WAF Identification)
[4] 📂 Fast Sensitive File Fuzzer (Probes .env, .git, backups, SQL dumps, APIs)
[5] 🧪 Targeted Vulnerability Tester (Directly audit SQLi, XSS, CORS, SSRF, LFI, Headers)
[6] 🧰 Security Tools Arsenal (26+ Tools across 10 Categories with 1-Click Auto-Install & Logs)
[7] 📄 View & Open Audit Reports (Browse, read, and 1-click open Cyberpunk HTML reports)
[8] 🤖 Interactive AI Security Copilot (Terminal chat, threat reasoning, and remediation code)
[9] 🌐 Start SentinelAI REST API Server (Launch FastAPI backend on localhost:8000)
[10] 🎯 Set / Change Active Target (Configure session memory target)
[0] 🚪 Exit

Just enter a number (1, 2, 3...). SentinelAI guides you through each step and remembers your active target so you rarely have to re-type anything!

3. Optional Direct Shorthand Commands

You can also run direct actions by passing arguments to sentinelai:

# Full Automated Scan with AI Remediation & Report Generation:
sentinelai scan https://example.com

# Launch Interactive AI Security Copilot:
sentinelai copilot

# Domain Reconnaissance & OSINT:
sentinelai recon example.com

# Sensitive File & Hidden Path Fuzzing:
sentinelai fuzz https://example.com

# Launch Security Tool Arsenal:
sentinelai tools

# Browse & open generated reports:
sentinelai reports

# Start REST API backend:
sentinelai api

4. Configuring AI Providers (Optional)

To enable live LLM reasoning, export any of the supported API keys:

# Groq (Ultra-fast Llama-3.3-70b - Recommended)
export GROQ_API_KEY="gsk_..."

# OpenAI
export OPENAI_API_KEY="sk-..."

# DeepSeek
export DEEPSEEK_API_KEY="sk-..."

# Venice AI
export VENICE_API_KEY="..."

Note: If no API keys are set, SentinelAI automatically uses its built-in offline Heuristic Rule Engine with zero configuration.

5. Running Tests

pytest -v

🌐 FastAPI REST API Endpoints

SentinelAI provides a REST API that powers the web dashboard and integrates into automated DevSecOps pipelines:

Method Endpoint Description
GET /api/v1/recon/{domain} Instant OSINT, DNS resolution, open port scanning, and WAF identification
GET /api/v1/tools Curated security tools inventory with local host installation detection
POST /api/v1/fuzz Fast sensitive path & file fuzzer with soft-404 wildcard filtering
POST /api/v1/quick-scan Asynchronous standalone scan with zero database dependencies
GET /api/v1/reports Index of all generated HTML, Markdown, and JSON audit reports
GET /api/v1/reports/{filename} Interactive view or download of generated security reports
POST /api/v1/scan Distributed cluster scan (uses PostgreSQL + Celery workers)
GET /api/v1/scan/{id}/status Real-time status polling for distributed scans

🐳 Optional: Distributed Docker & Web UI Mode

If you wish to run the full distributed cluster with PostgreSQL, Redis, Celery, and the React Web Dashboard:

cp .env.example .env
docker compose up -d

Access the web dashboard at: http://localhost:3000
FastAPI Backend documentation: http://localhost:8000/docs


⚖️ Legal & Ethical Disclaimer

SentinelAI is developed for authorized penetration testing, security auditing, vulnerability assessment, and educational research purposes only. Do not scan targets without prior written authorization from the system owners.


📄 License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.

About

SentinelAI: AI-powered web application security scanner for automated vulnerability assessment and penetration testing.

Topics

Resources

Stars

2 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages