The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to interact with Docker containers and check project statuses (e.g., docker exec, docker compose, curl). These are standard operations for managing the AMC and VIOS microservice stack and are used for deployment health checks and configuration.
[DATA_EXFILTRATION]: The skill handles RTSP URLs which may contain embedded credentials. It includes specific logic in scripts/run_rtsp_calibration.py to redact these credentials before printing or logging, mitigating the risk of accidental exposure. It also instructs the agent to use environment variables for sensitive tokens.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided RTSP URLs and camera names. While these are ingestion points for external data, the skill uses them strictly as parameters for structured API calls and does not interpolate them into prompts in a way that would influence agent behavior. Boundary markers are absent, but the capabilities are limited to specific REST API interactions.

amc-run-rtsp-calibration