Hey Product Hunt!
Excited to launch Strike48 Pick today, a lightweight, open-source network reconnaissance agent for security and IT teams.
The problem we built it for: Logs and remote scanning only tell you what's already been recorded. Unmanaged devices, rogue access points, open ports nobody documented, segments remote tools can't reach, none of that shows up until someone's physically inside the environment, which is usually a slow, manual process.
What Pick does: Pick handles port scanning, device enumeration, ARP mapping, mDNS/SSDP discovery, packet capture, and WiFi discovery, streaming results back in real time instead of batch exports. The single Rust codebase compiles to a desktop app, mobile app, terminal, or headless agent, so you can drop it wherever you need visibility. Once it's running, it It works standalone, or you can connect it to Strike48's agentic platform so findings get correlated with log data and investigated automatically.
Who it's for: Red teamers, pentesters, security researchers, and honestly anyone who's ever connected to conference Wi-Fi without thinking twice. I've been told a million times never to do that, and I've ignored the advice just as many times. When I finally pointed Pick at a conference network, two minutes in it flagged what looked like a rogue access point sitting quietly on the same network as everyone's laptop. Vendor demo or something worse, I still don't know. That's exactly the kind of blind spot Pick is built to catch.
It's free and open source, we'd love for you to try it, fork it, and tell us what to build next!
Happy to answer any questions in the comments!
Hey Product Hunt 👋
We're the team behind Strike48. We got tired of watching security folks lug around a bag of single-purpose recon tools for port scans, WiFi, device discovery and then stitch the output together by hand.
So we built Pick: one open-source agent, one Rust/Dioxus codebase, that compiles to desktop, mobile, terminal, or a headless drop-in. It wraps 90+ tools and runs them locally inside the environment you're actually testing, so you see the unmanaged devices, rogue APs and open ports that remote scanning and log-based detection miss. It's MIT licensed and free on GitHub—star it, fork it, break it.
For the pentesters and red-teamers: what's the one recon tool you wish behaved better across platforms? We're actively building and your wishlist genuinely shapes the roadmap.
The three-agent pipeline is a really nice touch for cutting through false positives. One thing that would make it even better is persistent state across runs, so the validator can compare today's results against yesterday's and flag new open ports or changed services as a diff rather than re-reporting the same baseline every time. Would save a lot of noise on long engagements.
@fahrettinnfsk Thanks for the kind words. The results have been really encouraging. We have a customer triaging over 46000 alerts a day with over at 97% pure AI closure rate. It is really exciting to see in action. Regarding a diff between runs for pick I think that is a great idea. That is again something that is accessible already through the Strikekit solution but putting it in Pick makes lots of sense. Again, we are dealing with an AI here so diffs between runs are a little less "pure delta" then we are used to from script runs.
Love that this is MIT licensed and runs everywhere from one codebase. One thing I'd really want is built-in collaboration, like sharing live scan sessions with teammates through a shared session link or token, so two people can watch and annotate findings as they happen instead of exporting and emailing reports back and forth.
@emredwp1 We are actively discussing that exact idea. Our goal is to get as wide a distribution as possible with is so enabling collaboration is definitely key. I can't promise yet when we will get the feature in but when I have clarity I will post back.
Would love to see a way to save and version recon sessions, like exporting a session config file that I can rerun on a different network or share with a teammate to reproduce findings exactly.
@trkankarakorzm very cool idea. We have a way to enable that through our Strikekit solution but baking it into Pick might make sense as well. At least at a basic level. Keep in mind though you are interacting with an AI so it isn't every exactly repeatable if you know what I mean.
Love that it runs from a single Rust codebase and bundles so many tools out of the box. One thing that would help a lot is persistent scan history per project, so you can diff results between runs and spot new open ports or devices popping up on a retest. Keep up the great work.
@neefz2n Thanks for the kind words, we have some really good people working on it. The diff scan idea is one we are 100% working on. Not sure on the ETA but will post when it drops. Remember though diffs are a little dangerous when using an AI because it isn't deterministic but it is definitely something we are pursing.