Kali Linux

MUI : A GUI Plugin For Binary Ninja To Interact And View The Progress Of Manticore

MUI (Manticore User Interface) project, we provide a graphical user interface plugin for Binary Ninja to allow users to easily interact with and view progress of the Manticore symbolic execution engine for analysis of smart contracts and native binaries.

ATTENTION This project is under active development and may be unstable or unusable. Please open an issue if you have any difficulties using the existing features. New feature development will be considered on a case by case basis.

Requirements

Aside from the Python requirements, we require the following:

  • Binary Ninja (latest development version) with GUI
  • git submodule update --init --recursive for Manticore submodule

Installation

MUI requires a copy of Binary Ninja with a GUI. Currently we are testing against the latest dev release(s) (2.4.2901-dev at time of writing).

Manticore only operates on native binaries within a Linux environment. EVM support has only been tested on Mac and Linux, and it requires the installation of ethersplay.

Python dependencies are currently managed using requirements.txt and requirements-dev.txt. You can run make init to set up a development environment.

  • Make the project available to Binary Ninja by creating a symbolic link to the plugins directory. From within the root of this repo, run the following:

For Mac
ln -s “$(pwd)/mui” “${HOME}/Library/Application Support/Binary Ninja/plugins/mui”
For Linux
ln -s “$(pwd)/mui” “${HOME}/.binaryninja/plugin
s/mui”

  • Make sure Binary Ninja knows about our Python virtual environment.
    • Open Binary Ninja’s “Preferences” -> “Settings” -> “Python” and ensure the “Python Interpreter” is correctly set to the Python path associated with the current virtual environment. Reference venv/pyvenv.cfg to find the base path.
    • Copy and paste the absolute path of the MUI project into Binary Ninja’s “Python Virtual Environment Site-Packages” and add the required /venv/lib/python3.<minor_version>/site-packages suffix for the site-packages path.
    • Restart Binary Ninja if necessary.

Development

Installing currently listed dependencies:

make init
For Mac (will be similar for Linux)
export PYTHONPATH=”/Applications/Binary Ninja.app/Contents/Resources/python:/Applications/Binary Ninja.app/Contents/Resources/python3″

Activating the python virtual environment (do this before running other make commands):

$ . venv/bin/activate

Code style and linting can be followed by running the following:

$ make format
$ make lint

Tests for code without Binary Ninja interaction can be run if you have a headless version of binary ninja available, otherwise only non-Binary Ninja tests will be run:

$ make test

Adding a new dependency can be done by editing requirements.txt or requirements-dev.txt and then running the following in the virtual environment:

$ pip install -r requirements-dev.txt -r requirements.txt

Usage (Native)

All MUI features can be accessed through either the right-click context menu or the command palette. Common features include:

  • Find Path to This Instruction / Remove Instruction from Find List
  • Avoid This Instruction / Remove Instruction from Avoid List
  • Add/Edit Custom Hook
  • Solve With Manticore / Stop Manticore

And the following widgets are available:

  • State List WidgetShows all the Active/Waiting/Complete/Errored states during manticore execution. Double-clicking a certain state navigates you to the current instruction of that state and renders its provenance tree in the graph widget.
  • State Graph WidgetShows the provenance tree for a certain state selected in the list widget. Tab can be used to expand/collapse the graph and double-clicking any of the state nodes navigates you to the current instruction of that state.
  • Run DialogThe run dialog is shown when you invoke the Solve with Manticore command. It allows you to configure the various manticore options, and the changes will be saved to the bndb file. Some example configs include using a combination of LD_PRELOAD and LD_LIBRARY_PATH environment variables to run the binary with custom glibc.
  • Custom Hook DialogThe custom hook dialog can be accessed using the Add/Edit Custom Hook command. It allows you to define a custom manticore hook at the selected address. You also have full access to the Binary Ninja API which allows you to add highlighting, comments, and more. A defined hook can be removed by setting the code input field to blank.

Usage (EVM)

EVM support is currently a bit limited. MUI EVM only supports the same feature set as the Manticore CLI tool. Available commands include:

  • Load Ethereum Contract
  • Solve With Manticore / Stop Manticore

And the following views are implemented:

  • EVM Run DialogThe run dialog is shown when you invoke the Solve with Manticore command. It allows you to configure the various manticore options, and the changes will be saved to the bndb file.
  • Run ReportThe report page shows the result of a manticore execution. It displays all the files produced using the Binary Ninja UI.
R K

Recent Posts

Install Pip on Ubuntu 18.04: Python 3 and Python 2 Setup Guide

Pip is the official package manager for Python and the standard way to install libraries from…

2 days ago

Install R on Ubuntu 18.04 from CRAN: Statistical Computing Setup

R is an open-source programming language and environment built for statistical computing and data visualization. It…

2 days ago

Install Jenkins on Ubuntu 18.04: CI/CD Server Setup Guide

Jenkins is an open-source automation server that makes it easy to build CI/CD pipelines. Continuous integration…

2 days ago

Install Android Studio on Ubuntu 18.04 with Snap and OpenJDK 8

Android Studio is the official IDE for Android development, built on JetBrains' IntelliJ IDEA platform. It…

2 days ago

Install and Configure GitLab on Ubuntu 18.04 with Omnibus

GitLab is a web-based, open-source Git repository manager written in Ruby. It includes built-in tools for…

2 days ago

Install Anaconda on Ubuntu 18.04: Python Data Science Setup Guide

Anaconda is the most widely used Python distribution for data science and machine learning. It bundles…

3 days ago