Control Chromecast

**Quick judgement**: This skill is a capable and fairly mature Chromecast controller built on top of `catt`. It offers robust media casting and playback control for devices on the same local network, including advanced options (subtitles, timestamps, yt-dlp options, state save/restore). It is best suited for users who already have Chromecast/Google Cast devices and a reasonably stable home/office network. **Strengths** - Broad functionality: device discovery, casting from many sources, playback and volume control, YouTube queue management, and state management. - Uses `catt` + `yt-dlp`, so it inherits support for a large range of media sites and formats. - Good fit for home automation and scripted media workflows where an AI agent orchestrates media playback. **Risks & limitations** - **Network dependency**: Requires Chromecast and host machine on the same network; mDNS issues or restrictive networks (corporate, guest Wi-Fi) can break discovery. - **External tool dependencies**: Relies on `catt` and `yt-dlp`; breaking API changes or site protections (e.g., YouTube) can cause intermittent failures. - **Local environment requirements**: For local file casting, specific TCP ports (45000–47000) must be open; misconfigured firewalls or NAT setups will cause casting failures. - **Content and volume control risks**: An automated agent could inadvertently cast unwanted or loud content to shared TVs/speakers; guardrails, user confirmation, or device whitelisting are advisable. - **Privacy considerations**: Casting arbitrary URLs or local files at the request of an AI agent may expose viewing habits or internal URLs to people in the vicinity; sensitive content should be handled carefully. **Recommended scenarios** - Home users with multiple Chromecast or Google Cast devices who want assistant-based control ("play this YouTube video on the living room TV"). - Smart home setups where an AI agent is allowed to start/stop media, adjust volume, or resume previous states as part of routines. - Office or studio environments where dashboards or web pages need to be cast to displays on demand. - Power users comfortable with `catt`/`yt-dlp` who want those capabilities exposed through an AI orchestration layer. Less ideal for highly locked-down networks, environments without Chromecasts, or scenarios requiring strict control over what content can be displayed without prior human confirmation.