Hardening
Baselines for operating systems, servers, browsers and cloud accounts. Every setting justified, every change tested rather than assumed.
Armory // Independent security lab
Armory is a lab for hardening systems, building defenses and testing them until they hold. What goes in stays protected: the doctrine, the tooling and the work behind them.
Security is mostly the discipline of saying no. These are the combinations every build in the lab answers to, from a laptop to a production edge.
Nothing is allowed until it is needed and named: ports, permissions, scripts, origins, people. Everything else stays shut, and the list of exceptions is short enough to read.
Every account, token and service gets the smallest key that does the job, for the shortest time it needs it. Standing admin access is a finding, not a convenience.
No single wall. Controls are layered so they fail independently, and one breach is a contained incident rather than the whole keep.
Plan for the day something gets in. Segment what matters, log what happens, and practise recovery before it is needed.
Identity, integrity, inputs and claims all get checked, including our own. A setting that has not been tested is an assumption.
Less code, fewer dependencies, fewer doors. What isn’t there can’t be exploited, so removing things is the cheapest fix there is.
The disciplines the lab keeps under lock: each one a layer, all of them aimed at systems that stay shut when someone leans on the door.
Baselines for operating systems, servers, browsers and cloud accounts. Every setting justified, every change tested rather than assumed.
Work out what is worth protecting, who would want it and how they would get in, before the first line of code.
Edge-first hosting, strict headers, least-privilege access, and secrets kept out of code and out of logs.
Logs that answer questions, alerts that mean something, and runbooks written before the bad day instead of during it.
Keys, certificates and vaults handled properly: scoped, rotated, backed up and recoverable.
Attack our own builds with misconfiguration hunts, fuzzing and red-team drills. Whatever breaks gets fixed and written up.
Questions, collaboration, or a vulnerability to report. Messages are sealed in transit and delivered straight to the lab’s inbox.