About

Most of what's on this site started as a question I couldn't stop turning over. How does a climbing rope actually absorb a fall? What's really happening when Windows blanks a protected window during a screen capture? Could I check a weight-based drug dose faster than flipping through a field guide? I build to answer those questions — and once the answer works, I clean it up enough that someone else can use it too.

I'm not precious about language or stack. Half of what I ship is a small Python library, half is a GUI someone can actually click, and every so often the interesting part is a Rust core doing the heavy lifting under a friendlier front end. The through-line isn't the tech; it's wanting to know how the thing works underneath.

A lot of my work traces back to pre-hospital and emergency medicine. Working weight-based drug math, triage scoring, and point-of-care decisions under time pressure makes you want tools that are fast, correct, and impossible to fumble. That's where drugdose, vitalscore, and Clinical_IQ came from, and it's the same instinct behind akribia — a computational model of how the brain weights evidence across conditions like autism, ADHD, and post-concussion syndrome.

When I'm not at a keyboard I'm usually outside — on rock, in the surf, or up in the mountains on skis. Climbing, surfing, and ski mountaineering all have a habit of turning into physics problems, and that's usually where the projects start. ropesim is a rope-dynamics engine that grew directly out of wanting to understand fall forces properly. The outdoors keeps seeding the backlog, too: freediving depth math, avalanche and snowpack notes, surf and tide calculators, and other small tools aimed at the places I actually go.

The other half of my time goes low-level. I like living close to the metal on Windows — DLL injection, tray daemons, clipboard internals, the undocumented corners of the API. capture-bypass is the flagship there. That systems work points toward where I want to head next: OT / XIoT security, where understanding how a system really behaves is the whole job. When I publish offensive tooling I try to do it the right way — clear, authorized-use framing, like kto — because the point is assessment, not bragging rights.

The pipeline is deep. Names are reserved and code is brewing for a batch of tools that map onto everything above — freediving depth math (barodepth), trail and tide utilities, more geospatial/OSINT work, and a handful of low-level Rust libraries. You can see the current shortlist in the In progress strip on the projects page; cards show up there the moment something ships.

Bigger picture, I'm aiming my systems work toward OT / XIoT security — the world where understanding exactly how a device behaves under the hood is the entire job. Expect the balance here to tilt further toward Rust, hardware, and security tooling over time, without losing the medical and outdoors threads that started it all.

github.com/Londopy · discord: _londo · back to projects