Builder Develops AI Agent Skill To Extract Slides And Notes From YouTube Videos Into Obsidian
This works really great with the agent /goal command, where a verification step (to only extract slides) is the finish criteria. I underestimated how hard this would be for an agent to do. But once I had the right workflow, I converted it into a skill. Now I just give it YouTube links of talks that I want to revisit or study at a later time.
This level of personalization wasn't possible before. And artifacts continue to really surprise me with how useful they are for research and learning.
And this can extend to all kinds of content and generate all kinds of artifacts on top of it, like code, scripts, animations, interactive knowledge checks, etc.
elvis@omarsar0Just built an insane new agent skill. It can perfectly extract slides from YT videos, then write notes, images, transcripts, and slides into Obsidian vaults. An HTML artifact allows me to navigate and add more notes as I listen. Should I release the skill?
By the way, a highly recommended talk by Anthropic on Memory and the new Dream feature. Lots of cool ideas there about the future memory in AI agents.
elvis@omarsar0This works really great with the agent /goal command, where a verification step (to only extract slides) is the finish criteria. I underestimated how hard this would be for an agent to do. But once I had the right workflow, I converted it into a skill. Now I just give it YouTube links of talks that I want to revisit or study at a later time. This level of personalization wasn't possible before. And artifacts continue to really surprise me with how useful they are for research and learning. And this can extend to all kinds of content and generate all kinds of artifacts on top of it, like code, scripts, animations, interactive knowledge checks, etc.