What problem does it solve? GCP VMs with ephemeral IPs get a new public IP on every reboot, breaking DDNS records and proxy subscriptions distributed to clients like Stash. This Skill automates the full recovery chain so subscriptions stay reachable without manual intervention. ## Core Features & Use Cases - Boot-triggered IP detection: A systemd oneshot service runs gcp-ip-check.sh at boot, compares the current IP against the last recorded one, and stays silent when nothing changed. - Full update pipeline: On IP change it updates Cloudflare DNS A records, regenerates the Clash-format override YAML, pushes it to a private GitHub repo, and restarts the local HTTP subscription server. - Local HTTP distribution: A Python http.server serves custom-sub.yaml and other subscription formats behind key-based authentication for Stash overrides. - Use Case: After a GCP VM restart, the new IP is detected automatically, DDNS and the GitHub-hosted subscription are refreshed, and Stash immediately pulls the working config from the local endpoint. ## Quick Start Ask the agent to simulate an IP change and run the gcp-ip-check.sh script to verify the DDNS update, GitHub push, and subscription server restart all work end to end.