Capabilities
We take the software apart and tell you exactly where it breaks.
The work spans the layers most teams never audit — kernels, system daemons, browser engines, and the network stacks underneath them. Below is what we actually do, in the order a finding usually travels.
Reverse engineering
Static and dynamic analysis of closed-source binaries — operating systems, privileged daemons, drivers, and the frameworks they lean on. We read disassembly, trace live processes, and diff releases to find what changed and what a change quietly broke.
Targets
Root-cause analysis
Once something misbehaves, we find out why — the exact instruction, the memory it touched, the invariant it violated. We separate the crash from the vulnerability and describe the fix in terms an engineer can act on, not just a repro that happens to fall over.
Advisory writing
A report a vendor can triage without a meeting: impact stated honestly, the minimal reproduction, affected versions, and the conditions the bug requires. Clear enough that the person patching it never has to email us to ask what we meant.
Coordinated disclosure
We manage the whole timeline — first contact, triage, fix verification, CVE assignment, and the public writeup — and we keep detail private until the patch is out. The goal is a fixed bug and a credited report, in that order.
Variant hunting
A fixed bug is a pattern. We take the shape of a patched issue and sweep for the same mistake elsewhere in the codebase, because the second instance is usually still open.
Have a target in mind?
Tell us what you're worried about. We'll tell you what we'd look at first.