https://github.com/jehna/humanify

Deobfuscate Javascript code using ChatGPT
https://github.com/jehna/humanify

Last synced: about 2 months ago
JSON representation

Deobfuscate Javascript code using ChatGPT

• Host: GitHub
• URL: https://github.com/jehna/humanify
• Owner: jehna
• License: mit
• Created: 2023-06-29T18:52:47.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2024-06-19T17:32:05.000Z (3 months ago)
• Last Synced: 2024-06-20T06:03:17.062Z (3 months ago)
• Language: TypeScript
• Homepage:
• Size: 273 KB
• Stars: 295
• Watchers: 10
• Forks: 19
• Open Issues: 10
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-ChatGPT-repositories - humanify - Un-minify Javascript code using ChatGPT (Others)

README

        
# Humanify

> Deobfuscate Javascript code using LLMs ("AI")

This tool uses large language modeles (like ChatGPT & llama2) and other tools to

deobfuscate, unminify, transpile, decompile and unpack Javascript code. Note

that LLMs don't perform any structural changes – they only provide hints to

rename variables and functions. The heavy lifting is done by Babel on AST level

to ensure code stays 1-1 equivalent.

### ➡️ Check out the [introduction blog post][blogpost] for in-depth explanation!

[blogpost]: https://thejunkland.com/blog/using-llms-to-reverse-javascript-minification

## Example

Given the following minified code:

```javascript

function a(e,t){var n=[];var r=e.length;var i=0;for(;i .env && npm start --  -o deobfuscated.js obfuscated-file.js

export OPENAI_TOKEN="your-token" && npm start --  -o deobfuscated.js obfuscated-file.js

OPENAI_TOKEN=your-token npm start --  -o deobfuscated.js obfuscated-file.js

npm start -- --key="your-token"  -o deobfuscated.js obfuscated-file.js

```

Use your preferred way to provide the API key. Use `npm start -- --help` to see

all available options.

### llama2

Prerequisites:

* You'll need to have a Python 3 environment with [conda installed][conda].

* You need a Huggingface account with access to [llama-2-7b-chat-hf

  model][llama2model]. Make sure to read the instructions on the model page

  about how to access the model.

[conda]: https://docs.conda.io/projects/conda/en/stable/user-guide/install/index.html

[llama2model]: https://huggingface.co/meta-llama/Llama-2-7b-chat-hf

Run the following command to install the required Python packages and activate the environment:

```shell

conda env create -f environment.yaml

conda activate humanify

```

You can now run the tool with:

```shell

npm start -- --local -o deobfuscated.js obfuscated-file.js

```

Note: this downloads ~13gb of model data to your computer on the first run.

## Features

The main features of the tool are:

* Uses ChatGPT functions/llama2 to get smart suggestions to rename variable and

  function names

* Uses custom and off-the-shelf Babel plugins to perform AST-level unmanging

* Uses Webcrack to unbundle Webpack bundles

## Contributing

If you'd like to contribute, please fork the repository and use a feature

branch. Pull requests are warmly welcome.

## Licensing

The code in this project is licensed under MIT license.