grok-pentest

KO | EN

gitlite — search

#python #hacktoberfest #typescript #llm #golang #go #android #macos #java #awesome #vue #awesome-list #security #openai

(★ 14)

Simple pentesting demo using Grok

grok-pentest 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Grok Pen Tester

**Real AI-Powered Penetration Testing Framework using Grok (xAI)**

This tool uses Grok to perform **actual** reconnaissance, directory enumeration, payload injection (XSS, SQLi probes, etc.), and analysis. It makes real HTTP requests and intelligently interprets results.

## ⚠️ Critical Legal & Ethical Warning

**This is a real penetration testing tool.** It actively probes targets with real requests and attack payloads.

- **YOU MUST HAVE EXPLICIT WRITTEN PERMISSION** before testing any target (including the local demo).
- Unauthorized scanning is illegal.
- The CLI will always prompt for confirmation before active testing.
- Use **only** on systems you own or are explicitly authorized to test.
- The authors are not responsible for misuse.

## Quick Testing (Recommended)

### Option 1: Local Vulnerable Demo (Best for testing)

We provide a deliberately vulnerable Flask app so you can test the pentester safely on your own machine.

**Terminal 1** — Start the vulnerable app:
```bash
python vuln_demo_app.py
```

If port `5000` is busy on macOS, use a different port:
```bash
PORT=5001 python vuln_demo_app.py
```

**Terminal 2** — Run the pentester against it:
```bash
grok-pentest scan http://localhost:5000 --type website
```

If you started on `5001`, scan that instead:
```bash
grok-pentest scan http://localhost:5001 --type website
```

The tool will discover multiple real issues (exposed admin/debug endpoints, missing security headers, XSS, information disclosure, etc.) because they were intentionally built in.

### Option 2: Public Test Target
```bash
grok-pentest scan https://httpbin.org --type website
```

---

## Installation

```bash
python3 -m venv venv
source venv/bin/activate
python -m pip install --upgrade pip
pip install -e .
export XAI_API_KEY="xai-..."   # Get your key from https://console.x.ai/
# Optional: force a specific model name if needed
export XAI_MODEL="grok-4.20-0309-reasoning"
```

If you already created your venv before this fix, reinstall dependencies:

```bash
pip install -e .
```

## Usage Examples

```bash
# Test the local vulnerable demo (recommended)
grok-pentest scan http://localhost:5000 --type website

# Test a public endpoint
grok-pentest scan https://httpbin.org --type website

# API-focused testing
grok-pentest scan https://httpbin.org/get --type api

# CLI fuzzing (example)
grok-pentest scan "echo" --type cli

# Dry run (no real requests or payloads)
grok-pentest scan http://localhost:5000 --type website --dry-run

# Export report
grok-pentest scan http://localhost:5000 --type website --output report.json
```

**Flags:**
- `--dry-run`: Simulate without making real network requests
- `--active` / `--passive`: Control whether active probing (dir enum + payloads) runs (active by default)
- `--key`: Override XAI_API_KEY

## What Happens During a Test

When you run the tool:

1. **Recon Phase** — Fetches page, analyzes headers, HTML, forms
2. **Active Probing** — Performs directory enumeration (`/admin`, `/debug`, `/api`, etc.) and sends Grok-generated attack payloads
3. **Grok Analysis** — Grok interprets responses, identifies issues, and generates prioritized findings + remediation steps
4. **Final Report** — Beautiful terminal summary + optional JSON report

You will see real findings because the demo app was built to be vulnerable.

## Features

- **Real active testing** (not just passive analysis)
- Directory enumeration with real requests
- Grok-generated attack payloads (XSS, injection probes, etc.)
- Intelligent result interpretation by Grok
- CLI fuzzing for command-line tools
- Rich terminal output with risk tables
- JSON report export
- Strong safety prompts and legal warnings

## Project Files

- `vuln_demo_app.py` — Intentionally vulnerable Flask app for local testing
- `grok_pentest/` — Main package (client, pentester, CLI)
- `demo.py` — Quick usage helper

## Development

```bash
# Install in editable mode
pip install -e .

# Run the demo instructions
python demo.py

# Start vulnerable test target
python vuln_demo_app.py
```

## Architecture

- `grok_pentest/client.py` — Grok/xAI API integration
- `grok_pentest/pentester.py` — Orchestration, recon, active probing, CLI fuzzing
- `grok_pentest/cli.py` — Typer CLI with safety gates and rich output

---

**Built with ❤️ by Grok. Use responsibly.**

**Reminder**: This tool is designed to help security researchers and developers test their own systems more effectively with AI. Always follow responsible disclosure practices.

## Features

### 1. **Website Pentesting** (Real)
- Fetches page + parses HTML (forms, inputs, links)
- Analyzes security headers (CSP, HSTS, etc.)
- **Directory enumeration** using AI-curated wordlist
- **Active payload testing** (XSS, SQL injection probes, open redirect, etc.)
- Grok interprets responses and identifies real issues

### 2. **API/Service Testing** (Real)
- Grok generates targeted attack payloads
- Tool sends them and analyzes responses for errors, leaks, status anomalies
- Tests for injection, auth bypass patterns, rate limits

### 3. **CLI Application Fuzzing** (Real)
- Grok generates creative edge-case inputs
- Executes the binary with them using subprocess
- Detects crashes, tracebacks, info leaks, timeouts

### 4. **Grok AI Brain**
- Generates intelligent test cases
- Analyzes raw responses for subtle issues humans might miss
- Produces prioritized findings + remediation steps
- Creates executive summary report

## Example Output

The tool outputs:
- Live progress with Rich formatting
- Grok's full reasoning and analysis
- Summary table of findings by risk level
- JSON report with all raw data

## Development

```bash
# Run tests/demo
python demo.py

# Install editable
pip install -e .
```

## Architecture

- `grok_pentest/client.py` — Grok API integration
- `grok_pentest/pentester.py` — Core orchestration + active testing logic
- `grok_pentest/cli.py` — User interface + safety gates

**Built with ❤️ by Grok for responsible security research.**

---
**Reminder**: This tool is meant to help security professionals and developers test *their own* systems more efficiently using AI. Always follow responsible disclosure and legal requirements.

grok-pentest

grok-pentest 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

grok-pentest

grok-pentest 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README