Marionette is a deterministic simulation testing (DST) library for Zig.

A tutorial about how to apply property-based testing to stateful systems.

A tool for comparing the design and analysis of surveys by simulating spatially-correlated populations

Simulation sandboxes to test AI agents against different APIs

Memory for AI agents - entity extraction, dual retrieval, and evolution tracking

IX-HapticSight: open safety authority for AI-controlled robots/XR. RGB-D perception→3D hazards→ALLOW/MODIFY/DENY physical action, with consent, force/tactile/proximity/thermal gating, ROS2 control, recovery, WebXR, replayable evidence & 183 tested checks. MIT + Responsible Use.

Deterministic crawl kernel in Rust — same seed, identical crawl, identical replay. BoringSSL TLS impersonation (JA3/JA4), 17-patch CDP stealth mode, BLAKE3 content-addressed storage, WARC++ format, distributed crawling, RAG extraction. 288 tests, 15 crates, 10K pages proven.

Idiomatic spec library for javascript

This repository contains the multi-physics validation models, mechanical-thermal calculations, and trajectory telemetry for a next-generation high-RPM internal combustion engine powertrain.

Testing the multivariate observer effects (MOE) framework

An infrastructure-to-vehicle (I2V) wireless speed-control system that dynamically restricts maximum vehicle speed in designated safety zones (school zones, municipal limits, construction areas) using Sub-GHz RF beacons and Drive-by-Wire throttle interception.

Raft Consensus

Find reproducible qualitative simulation transitions and turn them into pytest regressions.

Workflow orchestration engine written from scratch in Python, with leases, optimistic locking, and a deterministic fault simulator

Deterministic Rust testing utility for simulation and stochastic workflows

Deterministic simulation testing for TypeScript — find async races, deadlocks and lost retries, then reproduce them from a seed every time

Deterministic simulation testing for AI agents. Runs an agent inside a fake world where tools fail on purpose, and measures how often it claims to have done something it didn't — or does it twice.

Tamper-evident AI agent audit harness. Cryptographic witness chain, pre-call policy interception, and byte-stable replay. Prove what your agent did, tried to do, and was blocked from doing. Written in Rust.

Raft consensus in a seeded fault simulator: the five safety properties are checked after every step and any failure replays from its seed.