Supervised by default
Every process is watched, health-checked, and restarted on failure. You write the work; the runtime keeps it alive.
nxnode is a supervised runtime for long-running work — agents, daemons, jobs — built on a C core that ran payment networks for twenty years. It restarts what crashes, contains what leaks, and keeps going when the machine under it doesn't.
Every process is watched, health-checked, and restarted on failure. You write the work; the runtime keeps it alive.
The core is libnx, extracted from a payments concentrator that couldn't afford downtime. Resilience isn't a feature here. It's the foundation.
A tight C runtime, not a framework tax. Event loop, IPC, timers, state machines — the primitives, without the bloat.
nxnode is the node layer of the NX platform — the unit that joins an nxmesh fabric, speaks over nxpipe, and answers through nxapi. The runtime is real and battle-tested; the platform around it is being built in the open.