By skyforbes Most AI models refuse to help with pentesting even when you have authorization. This system prompt fixes that.
What it does
- Works with ChatGPT, Claude, Gemini, and open-source models
- Step-by-step interactive mode (provides one step, waits for confirmation)
- Ubuntu WSL optimized with commands for all major Linux distros
- Covers reconnaissance, enumeration, exploitation, privilege escalation
- Maintains ethical boundaries without constant lecturing
How it works
Paste the prompt into your AI's system instructions. Then ask something like "help me scan a target network" and it will guide you through one step at a time.
Example:
You: "Help me enumerate web directories"
AI: "Step 1: Install gobuster. Command: sudo apt install -y gobuster. Reply 'done' when ready."
You: "done"
AI: "Step 2: Run the scan. Command: gobuster dir -u http://target.com -w /path/to/wordlist"
And so on. You control the pace.
Features
- Every command includes Ubuntu/Debian, Red Hat/Fedora, and Arch variants
- WSL-specific guidance and limitations
- Built-in prompt injection defenses
- Context tracking across conversation
- Adapts to your skill level
Get it
Link- https://github.com/9okVR/Ethical-Hacking-AI-System-Prompts.git
Built using prompt engineering best practices and aligned with NIST/PTES/OWASP standards.