We're a small team that builds tools for people who are drowning in data, juggling too many interfaces, and wishing their infrastructure would just explain itself.
So we built a platform where AI agents talk to your systems, make sense of the noise, and actually do useful things. Then we open-sourced the pieces, because that's what you do.
Poke around. Break things. Build on top of what we've done.
How it works — the 30-second architecture tour
What's here — find the stuff you care about
Lab Notes — the latest from our lab notebook
Everything in Strike48 is built around hubs and connectors.
A connector is a standalone process that does one thing well — scan a network, manage Kubernetes clusters, run a C2 operation. It connects to a hub and waits for instructions.
A hub coordinates connectors. It's where you write workflows, chat with AI agents, and see what's happening across your infrastructure. Ask an agent "what's listening on this network?" and it knows to reach out to the right connector, assemble the results, and give you an answer.
The result: each tool stays small and focused, but together they form a platform where AI agents can actually do things in the real world.
AI-assisted engagements, network recon, built-in C2, and reporting that doesn't make you want to cry. From planning to deliverable.
Kubernetes cluster management and a unified desktop shell that puts all your tools in one window. Because alt-tabbing is not a workflow.
Autonomous agents, RAG-powered knowledge bases, and visual workflows. Point them at your data and let them figure it out.
Rust SDK, gRPC APIs, and a connector architecture that lets you plug anything into the platform. Your tool, our plumbing.
Pick a path and start exploring. Or just read the Lab Notes — we won't judge.
