ldrgen

ldrgen allows operators to generate shellcode loaders from a set of pre-defined templates and profiles with the aim of hopefully streamlining development and reducing the time it takes to get a new loader out the door.

Getting Started

There are available binaries on the releases page, or you can build from source for the latest version.

Usage

ldrgen profile --profile [path_to_profile] --shellcode [path_to_shellcode]

How does it work?

Source files (*.c) or loaders are saved as templates in the templates directory, placeholders are also supported to be templated during generation.
The template configuration is saved as config.yaml in the templates directory, this contains information about the loader, what source files it requires and whether any substitutions are required.
Profiles are saved in the profiles directory, these can be used and customized to your personal preference to utilize the templates you've created.

Source

These are your loaders, most of the time these are the artifacts that get caught by AV. Put some effort into making some of these in your own time, and you'll have a nice collection of loaders to use!

CreateThread_Xor.c is a good example of a simple loader that uses CreateThread to execute shellcode. It also has a xorShellcode function that will decrypt the shellcode before execution.

...

    xorShellcode( shellcode, shellcode_size, "${KEY}" );

...

Templates

The ${KEY} placeholder is explicitly defined in the config.yaml file, and is used to substitute the key used to decrypt the shellcode. This is a simple example of how to use placeholders in your templates.

  - token: "createthread_xor"
    key_required: true
    enc_type: "xor"
    method: "VirtualAlloc, xorShellcode, memcpy, CreateThread, WaitForSingleObject"
    files:
      - sourcePath: "Source/CreateThread_Xor.c"
        outputPath: "Main.c"
        substitutions: 
          key: "${KEY}"

      - sourcePath: "Source/Xor.c"
        outputPath: "Xor.c"
      
      - sourcePath: "Include/Xor.h"
        outputPath: "Xor.h"

      - sourcePath: "makefile"
        outputPath: "makefile"

Profiles

These are malleable configurations that can be used to generate a loader from a template.

{
    "name": "Default Loader, but with XOR (CreateThread)",
    "author": "@gatari",
    "description": "A simple shellcode loader that uses CreateThread with XOR encrypted shellcode.",
    "template": {
        "path": "/opt/tools/ldrgen/templates/config.yaml",
        "token": "CreateThread_Xor",
        "enc_type": "xor",
        "substitutions": {
            "key": "as@&(!L@J#JKsn"
        }
    },
    "arch": "x64",
    "compile": {
        "automatic": true,
        "make": "make",
        "gcc": {
            "x64": "x86_64-w64-mingw32-gcc",
            "x86": "i686-w64-mingw32-gcc"
        },
        "strip": {
            "i686": "strip",
            "x86_64": "strip"
        }
    },
    "output_dir": "./createthreadxor"
}

Releases

Standalone binaries are available, however templates and profiles will be ingested by ldrgen. The latest release should have these included.

Building from Source

cd ldrgen
make build

Design Principles

This tool is not designed to be turing complete by any means, it's designed to streamline the process of developing loaders- it is ultimately up to the operator to develop their own loaders and templates should they wish to do so.

ldrgen started as a personal project for encrypting beacon shellcode, then parsing it into a header file for use in a stageless loader. It was a pain to do this manually, so I decided to automate it- and it built from there.

Contributing

If you'd like to contribute, feel free to open a PR or an issue. I'm always open to suggestions and improvements. However, before you make an issue, do remember that the objective of this tool is to streamline the process of developing loaders, not to be provide a cheat code for bypassing AV.

thomasxm / ldrgen