HackDev AI

Enterprise-grade, AI-assisted web vulnerability scanner. HackDev AI combines a large library of purpose-built scan modules with autonomous agent orchestration and multi-LLM reasoning to plan, execute, and triage a security assessment end-to-end.

Part of the HackDev toolkit.

Highlights

• Autonomous orchestration — core/autonomous_orchestrator.py and core/coordinator.py plan and sequence scan phases, adapting to what earlier phases discover.
• 20+ specialist scanners — SQLi, XSS, XXE, SSRF, SSTI, IDOR, CSRF, CORS, JWT, GraphQL, RCE, deserialization, subdomain takeover, cache poisoning, request smuggling, and more (scanners/).
• Multi-LLM support — pluggable backends for Gemini, Groq, DeepSeek, and local LLMs (core/gemini_ai.py, core/groq_ai.py, core/deepseek_ai.py, core/local_llm_engine.py).
• Exploit chaining — core/ai_exploit_chain.py and core/exploit_executor.py chain findings into multi-step exploit attempts with AI-guided decision-making.
• Evasion-aware HTTP engine — Cloudflare bypass, TLS fingerprint spoofing, Tor routing, and browser automation (core/http_engine.py, core/tor_enforcer.py, core/browser_engine.py).
• Recon & intel — subdomain/API harvesting, CVE intelligence correlation, and tech fingerprinting (core/advanced_recon.py, core/cve_intelligence_manager.py, core/tech_detector.py).
• Reporting & compliance — structured findings, correlation, and compliance-oriented reports (core/reporter.py, core/compliance_reporter.py, core/vulnerability_correlator.py).
• Web UI — Flask-based dashboard for launching and monitoring scans (hackdev_ai.py, templates/).

Architecture

hackdev_ai.py         Flask entrypoint / web dashboard
core/                  Orchestration, AI engines, HTTP/browser engines, recon, exploitation, reporting
scanners/              Independent vulnerability-class scan modules
validators/            Deterministic result validation to reduce false positives
tests/                 Test suite

Requirements

• Python 3.10+
• See requirements.txt for the full dependency list (Flask, Playwright, cloudscraper, PySocks/stem for Tor, sslyze, and optional wrappers for nuclei, arjun, shodan, censys, etc.)
• API keys for whichever LLM backend(s) you enable (Gemini / Groq / DeepSeek), configured via .env

Setup

git clone https://github.com/raghubirrajmahato15/HackDev_AI.git
cd HackDev_AI
./install.sh
pip install -r requirements.txt
./start.sh

Or with Docker:

docker build -t hackdev-ai .
docker run -p 5000:5000 --env-file .env hackdev-ai

Configuration (API keys, scan defaults, wordlists) lives in config.py and is persisted under the reports/config directories created by ensure_dirs() on first run.

Usage

Start the service (./start.sh or python hackdev_ai.py) and open the dashboard to configure a target, choose which scanner modules and AI features to enable, and launch a scan. Findings, correlation, and generated reports are available from the dashboard and under the reports directory.

Legal & Ethical Use

HackDev AI is built for authorized security testing only — engagements you own, or where you have explicit written permission to test (pentests, bug bounty programs in scope, CTFs, your own lab environments). Do not point it at systems you do not have permission to test.

Status

Actively developed. Expect breaking changes between versions as scanner modules and the agent orchestration layer evolve.

License

MIT