Examples
Want to see what your camera can actually do? These are complete, real projects you can build, plus a list of ideas to spark your own. Every built project comes with full code on GitHub, so you can clone it, run it, and tweak it to fit your home.
Built projects
Each of these is a finished guide with working code, scaffolded by Claude Code from a plain-English README. Clone the repo or follow the README to get your own version running in minutes.
Daily Time-lapse Creator
Capture frames throughout the day and stitch them into an MP4 at midnight. Supports uploading to Google Drive or saving locally. No AI model required.
Motion Detection Event Recorder
Detect motion locally, record short clips on each event, and review captured footage in a simple browser dashboard.
Motion Descriptor
Detect motion locally and use an Ollama vision model to describe what triggered the event. Review descriptions and frames in a simple browser dashboard.
Person Spotted Notification
Detect motion, use an Ollama vision model to check for a person in frame, and post an alert image to Slack when one is found.
Kitchen Assistant
A conversational cooking assistant that uses the MCP take_snapshot tool to look at your worktop on request. Ask it to check whether food has browned, dough has risen, or ingredients are in order.
Ideas to build
Not sure where to start? Each idea below is written as a short prompt you can copy and paste straight into an AI agent like Claude Code, which will scaffold the project for you.