What problem does it solve? Running a desktop environment and background services on an RDK board wastes memory and CPU that could otherwise serve model inference. This Skill safely and reversibly turns an RDK device into a headless edge node by stopping and disabling lightdm and other non-essential services, with a one-command restore path. ## Core Features & Use Cases - Dry-run service audit: Lists candidate services (lightdm, cups, bluetooth, ModemManager, avahi-daemon) with their active/enabled state and planned actions as JSON, changing nothing by default. - Reversible apply and revert: --apply stops and disables candidates (with --now-only for temporary stops), and --revert re-enables and restarts everything. - Safety guardrails: The candidate list is hardcoded and never includes ssh or network management services, so the board cannot be cut off from remote access. - Use Case: Before deploying a vision inference pipeline on an RDK X5, run the dry-run, confirm the service list with the user, apply headless mode, and measure freed memory with a before/after memory audit. ## Quick Start Ask the agent to turn the RDK board into a headless inference node by disabling the desktop and unnecessary services, then confirm the candidate service list before applying.
