The Loki Framework
Overview
Loki is a tool designed to create sock puppet identities for OSINT (Open-Source Intelligence) investigators, private investigators, or anyone needing a fake online presence that cannot be traced back to their real identity. Setting up a sock puppet manually can be time-consuming and requires knowing where to source credible fake information. Loki automates this process, generating detailed identity profiles with ease.
Currently in its initial development phase, Loki provides essential details for sock puppet creation, such as names, contact information, and more. As of March 14, 2025, it also includes an AI-generated image from This Person Does Not Exist. The framework is actively evolving, with new features being added to enhance its capabilities.
Key Features
Technical Implementation
The Loki Framework is implemented in Python and leverages a modular command-line architecture to generate realistic fake identities for OSINT purposes. At its core, Loki uses structured data models to assemble personal, professional, and demographic details such as name, age, gender, nationality, occupation, and contact information. It integrates with online sources like This Person Does Not Exist to retrieve AI-generated profile images and supports output in multiple formats (TXT, JSON, CSV) for flexible data handling. Command-line arguments enable users to customize identity parameters, and the tool is designed to be platform-independent with minimal dependencies, making it portable across Linux, macOS, and Windows systems.
Challenges Overcome
We faced a few key problems. First, making fake identities look real but still fast to create. Second, getting profile images from the web without hitting errors. Third, keeping the tool working on Windows, macOS, and Linux. Small differences between systems made this tricky. Last, we wanted a command-line tool that felt simple, not confusing. Each fix made the tool better and easier to use.
Impact
Loki has grown fast. It has 75 stars on GitHub and good feedback from users. People like how easy it is to use and how real the generated profiles look. Some use it to test online setups, others for OSINT training. The project is open-source, so others can help improve it too. That keeps it moving forward.