WeChat API Fuzzer

This repository contains tools for analyzing and dynamically invoking WeChat Miniapp APIs.The goal is to evaluate which Android framework APIs are invoked by Miniapp APIs to create a mapping between them.

Overview

The main code is located in the Evalulate-WeChat folder. It leverages WeChat's devtools to execute miniapp APIs dynamically using the Chrome debugger protocol. Specifically, the code reverse engineers and customizes the WeChat debugging protocol to invoke APIs on targeted platforms.

Evalulate-WeChat

• Purpose: This is the core folder where the reverse-engineered WeChat debug protocol is implemented. It allows the evaluation of JavaScript miniapp APIs directly through the evaluate function in the debug protocol.
• Key Files:
  • evaluateParameters.js: Implements the functionality for dynamically invoking miniapp APIs using WeChat's debug protocol.
  • evaluateImproved.js, evaluateNew.js: Variations of the evaluation script for experimentation.
  • invariantApis.csv: List of APIs extracted using static analysis of WeChat apk.
  • undoc.json: JSON containing undocumented APIs (APIs extracted using static analysis but not mentioned in the documentation).

Eval-Output

• Purpose: This folder stores the evaluation results, showing which APIs successfully ran and which ones generated errors.
• Key Files:
  • accessdenied.csv, catch.csv, df_other.csv, other.csv: CSV files recording the API responses, such as denied access or errors caught during execution.

Fill-Parameters

• Purpose: This folder contains scripts to fill in the required parameters for miniapp APIs based on static analysis of the APIs. A template-based approach is used.

Usage

Evaluate Parameters:

The evaluateParameters.js file in the Evalulate-WeChat folder is the primary entry point. It feeds JavaScript code into the WeChat debug protocol and logs the output.

Filling Parameters:

The scripts in the Fill-Parameters folder take the WeChat APIs and their required parameters (from CSV files) and fill them using default parameters, following a templated approach.

API Evaluation:

After dynamically invoking the APIs, the results are stored in the Eval-Output folder. The CSV files record successful invocations and errors encountered during execution.

contributions welcome ! :)