Four guys with a background in computer science. We love to hack on challenging projects in our free time!
Basically we felt that the existing analysis tools were either not good enough, contained too many bugs or just didn’t suit our needs when looking into Android applications.
In the recent years the Android platform has gained a lot of momentum and this popularity lead to a wide range of applications and also a high amount of malware being produced.
When trying to figure out problems of applications and functionality of malware it is essential to have a good toolchain at hand that assists with the analysis. There were quite a variety of tools available and even the famous IDA Pro supports analyzing Dalvik bytecode by now (Dalvik is the interpreter that Android applications run in). However when using either of these tools we often experienced shortcomings and thought about improvements / extensions that often were difficult to accomplish even though mostly looking at open source software.
Thus we decided to start our own full blown solution which suits our needs, has more features and is more efficient than all these existing tools.
Two example use cases we focused on during the development are malware analysis and code auditing. This is why Dexter provides high-level software architecture information like package- or class dependencies while at the same time being capable of showing the raw bytecode as a control-flow graph.
Please contact us. Since Dexter shall serve information as comfortably as possible, any feature related requests/discussions are welcome!
None yet. However we want to implement a SDK which can run on the user’s client computer for further analysis. If you contact us about that, it might help to schedule our priorities ;)
We only perform static analysis techniques. Dynamic analysis might be a future feature if needed. If you contact us about that, it might help to schedule our priorities ;)
We are very limited in resources, especially time and money. If you increase one of them, we are likely to commit more code :)
The team consists of researchers developers from all kinds of areas and basically our own experience and knowledge led to the chosen architecture of Dexter.
We knew we wanted to support all platforms and with the current power of browsers went for a web-based front-end interface. The backend is implemented in Django but most of our actual analysis and the initial processing is done by components written in C, C++ and Python.
We will do everything possible to keep all comments / tags / modifications safe by doing backups and migrating database schema versions to newer versions when updating the software. However we can’t guarantee consistency a 100% as we are not a company and have limited resources for Dexter.
You can trust Dexter and the team that we know what we’re doing but occasionally something might go wrong. We are currently also working on a import / export feature for user-created data so that you could even make backups yourself.
Even though we are developing Dexter for quite a while now we are sure that we did not handle all cases and inputs correctly. This is why we host Dexter as a free public web-service - we want to improve our code and error handling by making all you guys work with the tool in ways that we might not have thought of.
The team does not have any “web developers” who know the quirks and incompatibilities of all browsers from the top of their heads. By using certain popular libraries and testing things in different browsers we tried our best to make it work for everyone. However once in a while we might break compatibility and we also did not take older browser versions into account.
Please understand that we will only support the major and most recent browsers out there with our small available manpower.
In principle yes. Please contact us for further information.