How to create a multifaceted PDF file?

Question

I like reading PoC problems || GTFO, and one thing I found great when I first discovered it was the "polyglot" nature of their PDF files.

Let's meet the explanation: if you consider, for example, their 8th release , you can unzip files from it; perform the encryption they talk about by running it as a script and even better (worse?) with their 9th problem, you can even play it as a music file!

Currently, I write small scripts every week, and every time I write a PDF file in LaTeX to explain these scenarios. Therefore, I would really like to create the same PDF files. Unfortunately, they explained (in part) in their first release how to include zip files, but they did it through three small thumbnails of cmd lines without any actual explanation.

So my question is basically: how can I create a polyglot PDF file containing things like zip, as well as a script shell that can be run using arguments like regular scripts?

I ask here about the process of creation, and not just about how this is possible. The ideal way for me would be that there are already some scripts or programs that make it easy to create such PDF files.

I tried to find the keywords "polyglot files" and other similar ones on the net and could not find any useful matches. Maybe this process has a different name?

I already read the presentation by Julia Wolf , which explains how everything works, but, unfortunately, I did not have time to apply the knowledge there in the real world, because I, unfortunately, was not used to playing with file headers and the way to create a PDF file.

EDIT: Ok, I read more and found the 7th edition of PoC || GTFO to be really informative about this. I can eventually create my own scripts to create such multifaceted PDF files, if I have some more time to review it.

Answer 1

I feel that parsing some file formats will be a good start. You can find many file format specifications for different file types through Google, but they can be tough, and you probably need some time to translate into any language that you use.

PDF: https://www.adobe.com/content/dam/acom/en/devnet/pdf/pdfs/PDF32000_2008.pdf

ELF: https://www.cs.cmu.edu/afs/cs/academic/class/15213-s00/doc/elf.pdf

ZIP: http://kat.sdf.org/zip_file_format.txt

The language you choose will require a way to read and write raw bytes (not just ascii alphanumeric), so maybe C will be good for more direct memory access. Some Python tricks can help with open scripting.

To analyze the files, you can create a tool like https://github.com/kvesel/zipbrk/ to separate them and then put them back together in a polygon format. For example, zip does not require that the section headers be at the beginning (or even adjacent in this respect), and the magic PDF number can be displayed in several places in the file. I also believe that I remember that the polyglot tool was included in one of the PoC publications || GTFO (possibly number 8 or 2) as a polyglot in a pdf file.

Do not forget the hacker bible! :) https://nostarch.com/gtfo

Answer 2

I played with the polyglot myself after attending the Ange negotiations, and also personally talked to him. You really need to understand file formats so that they can embed them into each other.
However, a short story, here are some links that I found extremely useful for creating polyglots:

Some older Google Code trunk
Polyclot material PoC

In particular, the second link (on github) will help you create polygons, but also understand how they work and how they are implemented. Since it is mainly Python material and very well written, it is very useful and easy to use.