Beacon Object File (BOF) for remote process injection, via thread hijacking, without spawning a remote thread. Accompanying blog can be found here. cThreadHijack works by injecting raw Beacon shellcode, generated via a user-supplied listener argument, into a remote process, defined by the user-supplied PID argument, via VirtualAllocEx and WriteProcessMemory. Then, instead of spawning a new remote thread via CreateRemoteThread or other APIs, cThreadHijack identifies the first enumerated thread in the target process, suspends it, and retrieves the contents of the thread’s CPU state via a CONTEXT structure. Then, the RIP register member of the CONTEXT
structure (on 64-bit systems) is manipulated to point to the address of the aforementioned remote Payload. Prior to execution, a routine is added to wrap the Beacon shellcode inside of a call to CreateThread – giving Beacon its own thread to work in, with this thread being locally spawned, versus being spawned remotely. The CreateThread routine is also wrapped in an NtContinue function call routine, allowing restoration of the previously hijacked thread without crashing the remote process. Beacon payloads for cThreadHijack are generated with a ‘thread’ exit function, allowing process continuation after the Beacon has been exited. Beacon listener names, when containing a space, must be placed in quotes.
On a Windows machine, open a x64 Native Tools Command Prompt for VS prompt. This can be done by pressing the Windows key and typing x64 Native Tools and selecting the prompt.
Change directory to C:\path\to\cThreadHijack.
nmake -f Makefile.msvc build
Load cThreadHijack.cna through the Cobalt StrikeScript Console with load /path/to/cThreadHijack.cna