Creating Turbo

CycleTLS Inspiration

Turbo is modern request library used to bypass cybersecurity checks. This project was first inspired by an already existing repo called CycleTLS, which was a NodeJS library that mimicked JA3 fingerprints to effectively do the same thing Turbo now does. However, I was not a fan of CycleTLS's object structure as it made using it in Dalphan AIO's development difficult. The main issue was that CycleTLS was made to be usable for someone who didn't necessarily know how JA3 fingerprints worked, which limited the customizability as the repo assumed a lot of design choices for you.

If you are interested with the technicals of the issue, I'll keep it short, but pretty much because you needed to control the fine details in Go, and the majority of the user facing code was written in NodeJS, CycleTLS had to find a way to merge both languages together while aiming to keep it simple. This led to having to create a request worker in NodeJS, however this worker had a lot of issues closing properly, and stalling on some requests.

Turbo's Creation

So here I am, faced with really only one option, to create my own library. I had previously attempted this a year ago, however found that it was too complicated for me to complete. I was definitely nervous this time around taking on this project, but I had no other choice so I figured I would brute force the knowledge and start work on Turbo.

I knew I had to use Go, as it already had a supporting library to assist in editing Go's net/http JA3 fingerprint. This library is Refraction Network's uTLS library. I had to learn Go as I had no prior experience with it, but I was able to get the hang of it pretty quickly. I was able to create a working prototype in 2 weeks, and after another month of testing and debugging, I was able to release a stable version Turbo.

Technical Details

The information I will share in this section is not meant to be a tutorial, but rather a high level overview of how Turbo works. If you are interested in learning more about Turbo, I would recommend checking out the repository.

JA3 fingerprints are a way to identify a client based on the TLS and Client Hello packet that is sent to a server. These JA3 fingerprints are composed of configurations including the ciphersuites, extensions, and elliptic curves that are used in the Client Hello packet. After collecting the JA3 fingerprints of popular modern browsers, I created configurations for those browsers that Turbo could use when making requests.

For these configurations to be used, Turbo had to actually expand upon the default net/http library. This was done by stepping through majority of the files in the net/http library, and adding a few lines of code to each file to allow for the JA3 fingerprint to be set. This was a very time consuming process, but it was necessary to ensure that Turbo could be used in any Go project without any issues.

For the last touch, Turbo had to also mimic the browser http header settings and order of these headers to ensure that the request would be as close to a real browser as possible. This was done by creating a map of the headers and their order, and then setting the headers in the same order as the map.