I got tired of 47 browser tabs. So I built BrowserSentinel.
Every online investigation starts the same way.
You open a tab for Shodan. Another for VirusTotal. Then WHOIS. Then social media. Then reverse image search… and before you know it, you’ve got 40 tabs open across three windows.
I worked in cloud investigations, and this was pretty much my daily setup. The tools were always there. What was missing was structure, a way to actually manage the work instead of just jumping between tabs.
So I built one. The idea is simple: bring everything into one place.
What it is
BrowserSentinel is a local-first OSINT investigation platform. It runs entirely on your machine as a desktop app. You do need to create any account or use any of the cloud service providers. ,Your data stays with you.
It combines case management with a launcher for 160+ OSINT tools, an embedded browser for investigations, some basic GEOINT capabilities, and AI support via Claude. Instead of juggling tabs, you work inside a single environment that’s built for investigations.
The embedded browser is probably the part I care about the most. It runs in its own isolated Chromium session, no shared cookies, no extensions, no history bleeding into your personal browsing.
When you find something useful, you can capture it in one click: screenshot, text preview, and a SHA-256 hash, all saved directly into your case.
The AI integration is connected to Claude through the Anthropic API. I’ve also added an MCP server, so Claude Desktop can read and annotate cases directly. In practice, that means you can ask it to look across multiple cases for shared indicators or patterns — and it actually helps instead of just summarizing.
How it’s built
Nothing fancy: Electron, vanilla JavaScript, and Node.js.
The whole app runs locally, and case data is saved directly to your disk. No backend, no external dependencies. That was a deliberate choice because your investigations shouldn’t depend on someone else’s infrastructure.
Availability
It’s free for personal use, learning, and research. If you want to use it commercially, there’s a paid license.
The code is on GitHub, and it will automatically convert to MIT in 2030.