Turn videos and papers into searchable Markdown libraries
They used OpenClaw to organize video and paper summaries into two separate Markdown-based libraries that could be searched later.
- What they did
- They ran OpenClaw on a Mac mini and manually set up youtube-summariser with pip plus Tobi’s QMD with bun, along with the package and LLM API keys. After that, they told Eve to learn both tools and use them to create a YouTube reference library. They also linked in a paper-fluff-cutter so it would save the important parts of papers as Markdown for QMD to index.
- How it went
- That produced two compact libraries: one for video material and one for papers. The YouTube library was able to handle questions well. No numbers or formal measurements were included.
- Tips
- Keeping the output in Markdown is what makes QMD’s indexing and search work smoothly. The setup work was done manually, so expect to handle installs yourself rather than relying on the agent.
Try it yourself with OpenClaw
I’d like you to help me make two searchable Markdown libraries, one for [videos] and one for [papers]. When I provide a [YouTube link or paper PDF/URL], extract the main points in [200-300] words, save them as a .md file in the correct folder, and refresh the index file with the title, date, and tags. Before you overwrite, remove, or submit anything, pause and ask me to confirm first, then tell me the saved file path once it’s done.
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Research web pages and drop screenshots into a document
A reviewer used Operator to gather web info and organize it in a Google Doc, but the agent struggled with browsing depth and could not handle screenshots.
Track public document changes for teaching updates
An educator used a Dot to watch public documents, rank likely teaching impacts, and compare draft notation with linked pages before suggesting edits.
Set up daily AI news briefs
A Grok bot monitored X and AI coverage, checked figures against sources, and delivered five polished research briefs each day.
Have Claude browse your own feed for post ideas
A writer used Claude in Chrome to scan a logged-in X feed and surface a few possible blog angles from posts with strong engagement.
Build a WHO data spreadsheet with separate tabs
A tester had ChatGPT Agent gather public WHO figures and sort them into a spreadsheet by indicator, country, and trend.
Gather recent LLM papers and turn them into a Word file
Nitika Sharma used Perplexity Comet to collect recent LLM and agent training papers, then had it format the results in a Word document.