🐍 Now Python 3.13 compatible !

🏄‍♂️ Description

ProtoDeep is an easy to use tool that allows you to decode and analyze protobuf data.

It is heavily based on the well-established Blackbox Protobuf package, and inspired by Protobuf Inspector for the custom definitions feature.
This project was originally intended to be integrated into the GHunt RDTK, but the script grew a lot and ended up becoming a tool that can help many people.

Features :

• CLI usage
• Python library usage
• Make your own definitions
• Easily match / filter data
• Export and compile on the fly

Example of using ProtoDeep on the Google's Play Store searchList endpoint, with custom definitions :

✔️ Requirements

• Python >= 3.10

⚙️ Installation

$ pip3 install pipx
$ pipx ensurepath
$ pipx install protodeep

It will automatically use venvs to avoid dependency conflicts with other projects.

💃 Usage

Help menu

Usage: main.py [-h] -t TYPE [-d DEFINITIONS] [-na] [-nk] [-s] [-b] [-hx] [-bi NUMBER] [-he] [-np] [-m MASK] [-mk MASK]
               [-mv MASK] [-f MASK] [-fk MASK] [-fv MASK] [-epf [PROTOFILE_FILENAME]] [-epd [PROTODEEP_FILENAME]]
               [-c [PYTHON_FILENAME]] [-n SCHEMA_NAME]
               [proto_file]

Positional Arguments:
  proto_file

Options:
  -h, --help            show this help message and exit
  -t, --type TYPE       Either protobuf (raw protobuf content), or protodeep (a ProtoDeep file).
  -d, --definitions DEFINITIONS
                        The file containing the custom protobuf definitions.
  -na, --no-autodetect  Don't try to autodetect if it's a raw HTTP request.
  -nk, --named-keychains
                        Show and extract only named keychains.
  -s, --stdin           Parse from stdin.
  -b, --base64          If this is a base64 input, so it automatically decodes it.
  -hx, --hex            If this is a hex input, so it automatically decodes it.
  -bi, --bruteforce-index NUMBER
                        The index up to which to try bruteforce to find Protobuf content. Default : 20
  -he, --hide-empty     Hide the empty values.
  -np, --no-print       Don't print the decoded protobuf.
  -m, --match MASK      Match anything with the given string. You can use '?' and '*' to wildcard match.Ex : "*token*"
  -mk, --match-keychain MASK
                        Match keychains with the given string.
  -mv, --match-value MASK
                        Match values with the given string.
  -f, --filter MASK     Filter anything with the given string. You can use '?' and '*' to wildcard match.
  -fk, --filter-keychain MASK
                        Filter keychains with the given string.
  -fv, --filter-value MASK
                        Filter values with the given string.
  -epf, --export-protofile [PROTOFILE_FILENAME]
                        Export the proto file with the definitions.
  -epd, --export-protodeep [PROTODEEP_FILENAME]
                        Export a protodeep file, to reuse in ProtoDeep.
  -c, --compile [PYTHON_FILENAME]
                        Compile protobuf into a Python file.
  -n, --name SCHEMA_NAME
                        Name of the schema when exporting into a proto file.

Concepts

Here are the main concepts to know when using ProtoDeep :

Output example

• Keychains : Since the protobuf is made of nested keys and values, keychains are a way to precisely identify a value in the decoded protobuf. It's the key sequence used to access the value.
• Pretty Keychains : This is the same as keychains, except that the keys are replaced by the names defined in the custom definitions.
• Type : Type of the value. Supported types are listed in this blackboxprotobuf's file.
• Value : The value found in the protobuf data.
• Iterator : ProtoDeep will try to autodetect repeated messages, and will print elements of these arrays with the i<position> key, so you can know the position of the element in the list.

Custom definitions file

• Definitions : It is a JSON file, containg a dict with the keychains as keys, and names as values. You can specify the type of a value by adding :<type> next to it. By doing so, ProtoDeep will detect it, and relaunch the decoding of the protobuf data with this new type. Note that it will only work when using protobuf data, not a protodeep file, since data has already been decoded.

Have fun 🥰💞

🧑‍💻 Developers

To use ProtoDeep as a lib, you can't use pipx because it uses a venv.
So you should install ProtoDeep with pip :

$ pip3 install protodeep

And now, you should be able to import protodeep in your projects like this :

from protodeep.lib import guess_schema

with open('protobuf_data.bin', 'rb') as f:
    raw = f.read()

protodeep_schema = guess_schema(data=raw)
protodeep_schema.pretty_print(hide_empty=True, filter_any=["*term_to_filter*"])
protodeep_schema.export_protodeep("obj.pdeep")

Testing

Thanks to learn-more, tests are now available, to test the CLI and lib usage !
You can launch the tests by doing :

$ pip3 install -r requirements-dev.txt
$ pytest

Tests are run automatically through GitHub Actions.

📕 Cheatsheet

Some examples so you know how to use protodeep :

Reading a protobuf file:

$ protodeep protobuf_data.bin -t protobuf

Read a protobuf file, provide a custom definitions file, hide the output, export to protofile & protodeep, and compile a Python file called "final.py" :

$ protodeep protobuf_data.bin -t protobuf -d search_ps_defs.json -np -epf -epd -c final.py

Names for the arguments --export-protofile / --export-protodeep / --compile are optional. If they aren't set, a default name will be used.

Read protobuf from stdin, provide a custom definitions file, match the keychain "11,1,1,2", hide the empty values, and filter lines where the word "access" and "denied" are present, and lines where the word "tiktok" is present:

$ curl -s <protobuf_endpoint> | protodeep --stdin -t protobuf -mk "11,1,1,2" -he -f "*access*denied*" -f "*tiktok*"

Matching / filtering arguments can be used as many times as you like.

Thanks

• The NCC Group for the super useful blackboxprotobuf project
• mildsunrise for protobuf-inspector
• The HideAndSec team 💜 (blog : https://hideandsec.sh)

Sponsors

Thanks to these awesome people for supporting me !

                

You like my work ?
Sponsor me on GitHub ! 🤗