Getting Started With Clearfront: Install It and Run Your First Scan

What is Clearfront?

Clearfront is an open-source OSINT tool that scans public data about a person, connects the findings, and has an AI security analyst write them up as an intelligence report. It runs on your own computer with your own API key, so nothing you scan leaves your device.

I built it to make a footprint self-audit take one command instead of an hour of manual searching. It ships 30 collection tools, a force-directed evidence graph, and four ways to run it: a command-line tool, an interactive prompt, a browser console, and an MCP server for Claude.

What you need before you install

• -Python 3.10 or newer
• -macOS, Linux, or Windows
• -An Anthropic API key for the AI analyst, or a local model with Ollama
• -A few optional command-line tools for deeper scans, which you can add later

You do not need any of the paid API keys to start. Many of the tools run keyless out of the box, so your first scan works with nothing but the Anthropic key.

How to install Clearfront

Clone the repository, install it with pip, and set your API key. Three commands.

git clone https://github.com/scottmartinanderson/clearfront.git
cd clearfront
pip install -e .
export ANTHROPIC_API_KEY=sk-ant-...

That installs the clearfront command. If you want the browser console, install the web extra as well:

pip install -e ".[web]"

Running your first scan

Type clearfront to start the interactive analyst, then give it an identifier like your own email or username. It decides which tools to run and reports back.

clearfront                         # start the interactive AI security analyst
clearfront email jane@example.com  # or run a one-shot lookup
clearfront username johndoe99
clearfront web                     # or open the browser console

Run it on yourself first. It is the fastest way to see what the tool does and the most useful scan you will run.

The browser console and one-click self-checks

Run clearfront web and it opens a local console at http://localhost:8080 with buttons that scan your own exposure without typing anything.

The console has four one-click checks: Check my exposure, Check my identity, Check my number, and Check my IP. Check my exposure is the one to start with. Everything runs locally and binds to your own machine by default, so the console is not exposed to your network.

Where your data goes

Nowhere. Clearfront runs on your machine with your keys. Your targets, your findings, and everything it collects stay on your device; I never see or store any of it. It works on public data only and does no intrusive probing, which keeps it on the right side of the acceptable-use policy.

Once it is running, the next thing worth learning is how to prompt the analyst well and how to pivot through the evidence graph. I covered that in how to prompt Clearfront AI security analyst.