Advanced Prompt Engineering & Bypass Framework
Created by: TΔreqShΔdØw xXx909999
PROMPT-DEMON is an elite, highly advanced command-line framework designed for generating complex prompt structures. It is built to push the boundaries of AI models by testing their alignment, safety filters, and operational limits.
By utilizing state-of-the-art obfuscation techniques, multi-layered theoretical bypass architectures, and linguistic fusion, PROMPT-DEMON crafts intricate prompts that evaluate how AI models handle extreme edge-case instructions, context manipulation, and role-play inception.
Flavor your prompt generation by selecting from various operational "modes", each injecting a different thematic persona into the generated prompt:
shadow: Stealthy, hidden commands.demon: Aggressive, authoritative overrides.void: Abstract, boundary-dissolving context.worm: Infiltrative, deeply embedded instructions.ghost: Untraceable, highly obfuscated commands.
- Intensity (1-10): Scales the complexity, visual dominance, and strictness of the prompt's instructions. Higher intensities demand absolute precision and cross-domain knowledge synthesis from the AI.
- Bypass Levels (
minimal,standard,maximum,apocalypse): Defines the depth of the bypass architecture applied. Theapocalypselevel deploys maximum unicode obfuscation and hidden encoded payloads.
- Unicode Mixing: Randomly replaces standard characters with visually similar characters from Latin, Greek, Cyrillic, Math, Fraktur, and Circled unicode blocks to evade basic keyword-based filters.
- Encoding Traps: Inserts Base64-encoded directives or warnings designed to confuse tokenizers.
Fuses and translates critical trigger phrases across multiple languages (e.g., French, German, Japanese, Arabic, Russian) within the same prompt to disorient language-specific safety alignment layers.
Automatically builds multi-layer prompt architectures using theoretical psychology and security concepts:
- Authority Framing: Assuming administrative or root privileges.
- Role-Play Inception: Placing the AI inside nested hypothetical scenarios.
- Reverse Psychology: Inverting expected instructions.
- Jailbreak Embedding: Injecting deep "developer mode" or "emergency protocol" directives.
Take any base prompt and automatically generate unique variants by applying semantic shifts, structure reversal, and varying levels of unicode obfuscation.
Paste any prompt to receive a calculated Bypass Score (0-100) based on:
- Prompt length and complexity.
- Unicode obfuscation ratio.
- Presence of authority markers.
- Number of embedded theoretical bypass techniques. Provides automated recommendations for improving the prompt.
- Python 3.x
- (Optional but Highly Recommended) Rich library for an enhanced, colorful CLI experience:
pip install rich
- Clone or download the repository.
- Navigate to the project directory:
cd path/to/prompt - Run the script:
python promptforge_10k.py
- Generate Single Prompt: Interactively create a highly customized advanced prompt.
- Generate Batch: Generate multiple prompts instantly with randomized, high-intensity parameters.
- Generate Variants: Mutate an existing base prompt.
- Analyze Prompt Effectiveness: Score a prompt's theoretical bypass potential.
- Custom Configuration: (Coming soon in v99.8).
- Export to File: Save your current session history to a
.txtfile. - View History: Review previously generated prompts during your session.
(Note: All generated prompts are automatically saved to generated_prompts.txt by default).
This tool is provided strictly for educational, security research, and academic purposes only. It is designed to assist AI safety researchers and developers in understanding Artificial Intelligence alignment, content filtering mechanisms, red-teaming, and the theoretical limits of prompt engineering. The creator assumes no responsibility for how this tool is utilized by end-users. Always adhere to the Terms of Service of the AI platforms you interact with.
TΔreqShΔdØw xXx909999 (ShadowHackr)
Feel free to reach out, follow my work, or contact me for collaborations:
- 🌍 Website: www.shadowhackr.com
- 💬 WhatsApp: Contact Me Directly
- 📘 Facebook: ShadowHackr
- 📸 Instagram: @shadowhackr
-
- 📸 art: @shadowhackr