thebabush/llvm-jutsu
Anti-LLM obfuscation via finger counting
This project helps software developers protect their proprietary software by making it harder for AI models to reverse engineer critical parts of their code. It transforms sensitive numerical constants within compiled programs into visual hand gestures. The output is a binary where these constants are hidden as images, making it difficult for AI to automatically extract information, while still allowing a human to decode them manually.
182 stars.
Use this if you are a software developer creating a license check, digital rights management, or other sensitive code and want to prevent large language models from easily extracting secret values from your compiled binaries.
Not ideal if your primary concern is high-performance code or minimizing binary size, as this technique adds significant overhead and increases file size.
Stars
182
Forks
10
Language
C++
License
MIT
Category
Last pushed
Dec 22, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/thebabush/llvm-jutsu"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
google/oss-fuzz-gen
LLM powered fuzzing via OSS-Fuzz.
cyberark/FuzzyAI
A powerful tool for automated LLM fuzzing. It is designed to help developers and security...
prompt-security/ps-fuzz
Make your GenAI Apps Safe & Secure :rocket: Test & harden your system prompt
fuzz4all/fuzz4all
🌌️Fuzz4All: Universal Fuzzing with Large Language Models
ChatAFLndss/ChatAFL
Large Language Model guided Protocol Fuzzing (NDSS'24)