InjectIQ v2

AI-Native Database Exploitation Framework

InjectIQ is an AI-native, autonomous database exploitation framework designed for authorized penetration testing and red team engagements. It unifies SQL, NoSQL, and GraphQL injection techniques with real-time AI analysis via Ollama to dynamically adapt payloads, bypass modern WAFs, and achieve full data extraction with minimal manual intervention.

Disclaimer: This tool is intended for authorized security testing only. Use against systems you own or have explicit written permission to test. Unauthorized access to computer systems is illegal.

---

Features

• Multi-Database Injection Engine

  • SQL: MySQL, PostgreSQL, MSSQL, Oracle, SQLite, MariaDB
  • NoSQL: MongoDB, Redis, Cassandra, DynamoDB, Elasticsearch
  • GraphQL: Apollo, Hasura, and generic endpoints

• AI Copilot (Ollama Integration)

  • Real-time WAF response analysis and custom tamper generation
  • False-positive elimination via pattern recognition
  • Novel bypass technique generation for unknown WAFs
  • Autonomous decision-making for extraction strategy

• WAF & CDN Bypass

  • Cloudflare, Akamai, Imperva, AWS WAF, F5 ASM, Fortinet
  • CDN origin discovery and direct backend targeting
  • Dynamic tamper script selection per WAF fingerprint

• Advanced Techniques

  • HTTP Request Smuggling (HRS) for WAF bypass
  • Parameterized Query Bypass
  • Second-Order Injection
  • Time-based, Error-based, Union-based, Boolean-based, Stacked queries
  • Out-of-band (OOB) extraction via DNS/HTTP

• Autonomous Operation

  • Full pipeline: Probe → Fingerprint → Bypass → Extract → Report
  • Minimal configuration required for standard engagements

---

Architecture

┌─────────────┐    ┌─────────────┐    ┌─────────────┐    ┌─────────────┐
│   Probe     │───▶│  Payload    │───▶│   Inject    │───▶│  Extract    │
│   Engine    │    │   Engine    │    │   Engine    │    │   Engine    │
└─────────────┘    └─────────────┘    └─────────────┘    └─────────────┘
       │                   │                  │
       ▼                   ▼                  ▼
┌─────────────────────────────────────────────────────────────────────┐
│                         AI Copilot (Ollama)                          │
│  WAF Analysis • Tamper Generation • Decision Logic • False-Pos      │
└─────────────────────────────────────────────────────────────────────┘

Module	Purpose
probe.py	Endpoint discovery, WAF fingerprinting, DBMS identification, CDN origin bypass
payload.py	Payload generation per technique per DBMS, tamper engine
inject.py	Orchestrates the full injection loop with AI feedback
comparator.py	Response differential analysis, true/false determination
smuggling.py	HTTP Request Smuggling (TE.CL, CL.TE) attacks
param_bypass.py	Parameterized query bypass and second-order injection
ai_copilot.py	Ollama integration for real-time AI analysis
cli.py	Click-based CLI interface

---

Installation

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

Optional: AI Copilot Install Ollama and pull a model:

ollama pull qwen2.5-coder:7b

---

Usage

Quick Scan

python -m injectiq scan --url "https://target.com/page?id=1"

GraphQL Target

python -m injectiq scan --url "https://target.com/graphql" --graphql

POST Body Injection

python -m injectiq scan --url "https://target.com/api" \
    --method POST \
    --data '{"q":"test"}'

HTTP Request Smuggling

python -m injectiq smuggle --url "https://target.com"

Parameterized Query Bypass

python -m injectiq bypass --url "https://target.com/page?id=1"

Second-Order Injection

python -m injectiq second-order \
    --store-url "https://target.com/register" \
    --trigger-url "https://target.com/admin"

Full Autonomous Dump

python -m injectiq dump --url "https://target.com/page?id=1" --dbms mysql

---

Requirements

• Python 3.10+
• httpx>=0.27.0
• click>=8.1.0
• Ollama (optional, for AI copilot)

---

Training Data

• sqlmap_tamper_training.json — Historical WAF bypass payloads for ML training

---

License

MIT License — See LICENSE for details.

---

Author

Hamza Zaka Khan — github.com/hamzazakakhan