ZOT is a digital immune system implemented as a single Rust file that detects anomalies in a system's physical substrate. It operates without machine learning or training data, instead employing a Darwinian kinetic proofreading approach inspired by biological immunology. The tool calibrates itself against the system it runs on, evolving 500 immune receptors through a process of negative selection, and uses three independent physical probes to sense the environment in real time.
The anomaly detection mechanism relies on three distinct sensory channels: memory, clock, and allocator. The memory probe performs a 1 MB pointer chase across cache lines to sense cache hierarchy latency, repeating this process six times per reading. The clock probe executes 80 consecutive Instant::now() calls to measure timer access contention under load. The allocation probe cycles through 64 KB allocations, touches, and drops to detect memory subsystem stress. Each probe takes 50 readings per cycle, with the median used to define the system's state at each instant.
ZOT's architecture maps its signal processing to immunological concepts. Thymic selection tests 2,000 candidate receptors against 60 seconds of quiet system activity, eliminating those that react to the normal state and retaining 500 survivors as the immune repertoire. Kinetic proofreading ensures that only persistent, above-threshold signals trigger a response, filtering out transient noise. The population of receptors evolves continuously: every five cycles, low-confidence receptors are removed, top performers are cloned, and their offspring are mutated, allowing the system to adapt in real time. Regulatory suppression mechanisms adjust sensitivity during recovery phases to minimize false positives.
The tool has no external dependencies and is delivered as a single Rust source file. It can be run by cloning the repository, building with Cargo, and executing the binary, with an optional Python script for analysis. ZOT achieves accuracy rates between 93% and 95%, with false positive rates ranging from 0% to 2%.
ZOT sits in PulseGate's Other infrastructure category. It focuses on detecting hardware and system anomalies efficiently without relying on machine learning or external dependencies. ZOT is an open-source project aimed at system administrators and developers. The project is open source (MIT). ZOT is available on the command line.
It is developed by ZOT Research, and the product first shipped in 2026. Across PulseGate's embedding index, ZOT has few near neighbours, marking it as relatively distinct. Among its 5 catalogued features are anomaly detection, no dependencies, and rust implementation.
Latest indexed changes and source events
Other apps tracked under the same category.
