robotics-security

Open authorization and accountability infrastructure for physical AI: short-lived capabilities, local action gates, replay protection, signed receipts, post-quantum ML-DSA-65, fleet policy/revocation distribution, and a zero-dependency offline browser verifier.

Awesome curated list of ROS / ROS 2 & robotics security - pentest tools, checklists, CVEs, SROS2, DDS, research. Maintained by IOTSRG.

Hands-on CAE-CD cybersecurity lab & defense-in-depth guardrail sandbox for LLM prompt injection, indirect injection, FERPA privacy, and autonomous quadruped robotics security.

ROS 2 application-layer secure command pipeline using ML-DSA-44 post-quantum signatures, replay protection, TTL validation, and benchmark visualization.

Secure, quantum-resistant coordination and accountability layer for multi-robot and Physical AI agent fleets.

Sanitized, byte-verified evidence for reproduced CVEs in ROS 2 / robotics middleware (Fast DDS, DDS Security).

NEUROSIS Research — independent AI security research on autonomous agents, memory, multi-agent systems, VLM and embodied AI.

A node-spoofing attack on a ROS 2 command topic and a tested token-authentication defence against forgery, key-guessing and replay. Personal continuation of my MSc thesis.