AI Assistant

Send one sentence to your AI assistant (OpenClaw, etc.) and it can take over the cameras — view the feed, query events, pan/tilt, and speak through the camera.

How to use it

Send the paragraph below to your AI assistant; it will download and install the integration package on its own and guide you through configuration:

text

Download and extract https://cdn.yun-kan.com/yunkan-skill-latest.tar.gz, read the SKILL.md and README.md inside, install this SkyView integration package as instructed, then guide me to generate an API Token and complete the connection setup.

What you need to prepare

An AI assistant that can run command-line — OpenClaw, Claude Code, Hermes, etc. all work
An API Token — open the web admin at http://serverIP:23406 in a browser, log in, go to Settings → API Token → New Token, and copy the generated string starting with skv_pat_; the AI asks for it during setup

An API Token equals the admin password

Whoever has it can view all your cameras, pan/tilt, and make announcements. Don't post it in chat groups / screenshots / public repos. If you suspect a leak, revoke that Token in the admin and regenerate it.

You can talk to the AI like this

Once set up, just speak plainly to the AI assistant, for example:

"Show me what the living-room camera looks like right now" — the AI pulls a live frame and describes it to you
"What events happened in the backyard yesterday" — the AI queries the event log and summarizes
"Turn the garage camera to the left" — the AI controls the pan/tilt
"Announce through the front-door camera: please leave the package at the door, thanks" — the AI plays it through the camera speaker
"Check the system status and how much storage is left" — the AI pulls the system overview