maigret

Give it a handle, get back a dossier: which sites hold an account under that name, what the profile pages say, and where the trail of related usernames leads next.

The tool checks for accounts across a very large number of sites, over three thousand by its own count, and gathers whatever information the profile pages and site APIs expose, all without API keys. It is Python under MIT, with a homepage at maigret.app and around 36,800 stars on the repo.

You do not have to hit all three thousand sites at once. A default run covers the 500 highest ranked sites by traffic, which keeps things fast and relatively polite. There is an option to scan everything, and tags narrow the sweep by category or country. Because the tool is embeddable as a Python library, the same machinery can live inside another project, and searches recurse through discovered usernames to pull related handles into the net.

Sites block scrapers, so Maigret detects and partially bypasses blocks, censorship, and CAPTCHA. The site database refreshes itself from GitHub once per 24 hour run, with a locally stored copy as fallback when the fetch fails. That self updating design keeps the target list current without asking anything of the user.

On Windows there is a double click path: an executable that asks for a username and runs the default search. The terminal exposes the extra options, and Docker images exist for both CLI mode and a web interface mode. A sponsor section in the README lists residential proxy services, which tells you something about who the heavy users are.

As a research tool it is effective and slightly unsettling in equal measure. The ethics live with the person running it.

• Source 1 · Project repository
• Source 2 · README