6 projects
agentack
Test whether human approval controls for AI agents actually work.
autonomyfit
Evidence-aware model selection, deployment validation and local benchmarking for edge AI and autonomous systems.
kinemica-verify
Open-source verification infrastructure for physical-world work.
worldstate-check
Deterministic postcondition verification for AI agents and autonomous systems.
agent-boundary-check
Verify what AI coding agents can actually read, write, execute and reach using synthetic canaries.
rosbag-doctor
Check ROS 2 bags for timing gaps, bad rates, timestamp problems, coverage loss, and sensor-sync issues.
