Most ROS2 tutorials skip the security chapter, which is why so many research robots ship with unauthenticated DDS topics that let anyone on the network publish to /cmd_vel. This skill walks you through SROS2 keystore generation, per-node certificate setup, and writing the governance and permissions XML that actually enforces publish/subscribe access control at the DDS layer. It covers the full attack surface from open multicast discovery to unsigned firmware updates, and includes practical launch file snippets for running nodes in Enforce mode. The physical-cyber safety framing is honest: on a robot, a compromised topic isn't just a data leak, it's a collision or worse.
npx -y skills add arpitg1304/robotics-agent-skills --skill robotics-security --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
hoodini/ai-agents-skills
agamm/claude-code-owasp
addyosmani/agent-skills
giuseppe-trisciuoglio/developer-kit