“How do you know that?” Why I built Furgaris (BrowserSentinel second part)
If you work in investigations, fraud or threat intelligence, you know these two questions very well.
The first one: “What’s the status of the investigation?”
The second one usually comes at the end of a case, from a manager, a colleague, a lawyer or a client: “How do you know that?”
And very often the honest answer is somewhere between forty browser tabs, a notes file you’ll never find again, and your memory.
The tools were never the problem. WHOIS, VirusTotal, Shodan, a reverse image search: they all work fine. What I missed in some of the teams i worked was structure. One place where the case lives, where every step is recorded, and where I can show how I reached a conclusion. So I started building one to put all my frustrations, all the teims that a dev team told me they couldnt (either becasue my team was too small or their lack bandwith or resources). I called this tool: Furgaris.
Where the name comes from and what the tool does
Furgar is a Catalan word. It means to search with intention, to dig around because you’re curious and want to find something. It felt right for a tool like this.
Furgaris puts case management and OSINT in one place:
Cases. Open a case, set the priority and TLP, assign an analyst, and track tasks, notes, a timeline and evidence.
An audit log. Every action is timestamped, so “how do you know that?” has an answer.
Overlaps between cases. When the same IP, email or domain shows up in another case, you see it as soon as you open the case. No need to write SQL queries to find out whether two investigations are connected.
OSINT in one place. OSINT searches, Google dorks, Shodan and Censys, analyst tools (encode/decode, hashes, defanging IPs before they go into a report), an entity canvas, image tools and a URL watchlist.
Team view. A dashboard with what’s open, what’s past due and who has bandwidth, plus a 12-week view with money lost and prevented.
Optional AI. Threat-analysis support, an intel search agent and AI-drafted reports, using your own Claude API key. AI reports can also run locally with Ollama, so nothing leaves your machine. Anything the AI writes is a draft for you to check, never a finding.
This is a proof of concept
I want to be clear about this. Furgaris works, but it’s far from finished. Some parts are rough, some tools you use every day won’t be there, and some things I built might not be useful at all.
That’s exactly why I’m sharing it now. I built it from my own experience, but I don’t want to keep building it only from my own point of view. I’d rather hear from people who do this work every day.
How you can help
Watch the walkthrough. It’s 24 minutes and shows every part of the tool.
Try it. Download it from GitHub: github.com/Bertosk. There’s a Windows installer, or you can open the HTML version in Chrome or Edge. The user guide is in English, Catalan and Spanish.
Tell me what you think, good or bad. What works, what gets in your way, what’s missing, and what I should remove. Open an issue on GitHub or email me at albertplanas@albertplanas.eu.
Last Note
One request: please never share real case data when you report something. GitHub issues are public.
And if OSINT and case management is your thing and you’d like to be more involved, as a tester, with ideas, or more than that, please reach out. I’d be happy to talk.
Licence
Furgaris is free for personal use, learning, academic research and teaching, and non-profit journalism. Commercial use (companies, public bodies, law enforcement) needs a licence, so get in touch first.
Thanks for reading, and thanks in advance to anyone who gives it a try!