Scans 840+ online platforms using machine learning and heuristics to discover and correlate social media profiles.
Find profiles across 840+ sites, extract metadata, and correlate digital identities without manual searches.
Aliens_eye is a Python-based OSINT scanner designed to locate social media accounts across 840+ platforms using machine learning and structural signals. Content creators, researchers, and digital investigators can use it to map online footprints and detect brand impersonation. It replaces naive status-code checks with ML detection to filter out false hits.
The tool converts web responses into 30-dimensional feature vectors to determine if a target username exists on a given network. It extracts metadata such as avatars, bios, and display names, then clusters accounts that share image hashes or links. Scans run from the command line, an interactive terminal UI, or an MCP server for LLM agent integration.
This tool suits brand safety analysts, digital investigators, and audience researchers needing to verify online presence across hundreds of domains. It is not suitable for users looking for a web interface or a zero-code solution. Effective use requires familiarity with command-line tools and proxy setups.
Blends logistic regression with 30 DOM and HTTP signals to reduce false positive account detections.
Clusters found profiles into target identities by matching avatar image hashes, bios, and shared links.
Extracts additional handles mentioned inside discovered bios and automatically re-scans them.
Re-scans targets on set intervals and sends newly discovered accounts or profile updates to webhooks.
Exposes scanning capabilities directly to LLM agents using an integrated MCP server mode.
Aliens_eye is free and open-source software under the MIT license. There is no hosted paid version, so your costs depend on your own proxy bandwidth and compute infrastructure.
880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
It tests hits against a control username that cannot exist and uses a trained machine learning model alongside 30 structural indicators instead of relying solely on HTTP status codes.
Yes, installing the optional `[serve]` extra lets you run `aliens_eye serve` as a Model Context Protocol (MCP) server for LLM workflow integrations.
See Zekai first in Google