hamzafarooq What this skill does
When triggered, ask the user:
- What topic or question they want to research
- How many videos they want (default: 5)
- Any preferences — e.g. recent only, beginner-friendly, a specific channel
Then use Brave MCP to navigate YouTube, search for the topic, and return a structured list of relevant videos.
How to execute
- Navigate to
https://www.youtube.com/results?search_query=TOPIC— replace TOPIC with the user's input (URL-encode spaces as+) - Take a snapshot to read the search results page
- Collect the top results — for each video note: title, channel name, view count, publish date, and URL
- If results look off-topic or thin, refine the search and try once more
- Close the browser when done
Output format
Produce this write-up directly — no preamble:
YouTube Research: [Topic]
Search query used: [exact query]
| # | Title | Channel | Views | Published | Link |
|---|---|---|---|---|---|
| 1 | ... | ... | ... | ... | Watch |
| 2 | ... | ... | ... | ... | Watch |
| ... |
Top pick
[Video title] — One sentence on why this is the best starting point for a PM or non-technical person.
What the results tell you
2-3 sentences on the overall landscape: Is this topic well-covered? Are results mostly technical or beginner-friendly? Any dominant voices or channels worth following?
Suggested follow-up searches
2-3 alternative queries to go deeper on subtopics.
Rules
- Use plain English throughout
- If YouTube blocks the search or results are empty, tell the user and suggest they try the query manually
- Do not fabricate view counts or dates — only report what you read from the page
- After the write-up, ask: "Want me to save this to docs/youtube-research.md?"