Open-Source Intelligence (OSINT): Turn One Username into a Full Digital Dossier with Maigret

What it is

Maigret (github.com/soxoj/maigret) is a Python username-OSINT tool. The page is a Kali walkthrough of installing it and running the main flags. Maigret checks the 500 most popular sites by default, ordered by popularity. Sherlock inspired it, and the page positions Maigret as the more thorough option.

What to do

Install into a virtualenv:

git clone https://github.com/soxoj/maigret && cd maigret
python3 -m venv .venv
source .venv/bin/activate
pip3 install .
maigret --help

Core invocations as the page gives them. The page's typography renders double hyphens as en-dashes, so the flags below use --:

maigret username                      # top 500 sites
maigret username -a                   # full site database
maigret username --site GitHub        # one site
maigret username --tags coding        # filter by site tag
maigret --permute firstname lastname  # generate and search username variants
maigret username --html               # standalone HTML report with images
maigret username --ai                 # LLM-written summary

The page also says CSV, TXT, JSON and PDF reports and a graph flag exist, but it does not show their flags. Check --help for those.

For --ai, create ~/.maigret/settings.json. It points at OpenRouter with openai/gpt-4o-mini:

{
 "openai_api_key": "sk-or-v1-YOUR_OPENROUTER_KEY",
 "openai_api_base_url": "https://openrouter.ai/api/v1",
 "openai_model": "openai/gpt-4o-mini"
}

Caveats

• The --ai mode sends findings to an external LLM provider, and the API key sits in a plaintext settings file. Consider this before running it on a real target.
• The page is a basic tour. It gives no accuracy or false-positive figures and no rate-limit or detection behavior. It also gives no guidance on authorization or scope for investigating a person.
• The comparison with Sherlock is only a one-line contrast, with no benchmark.