AI Agent Use Cases

Added today

44All 189

Prompt to Prepare a Shopping Cart from a Recipe Photo

The author reports that Grok Bot receives a recipe photo, organizes the ingredients and quantities, and creates a shopping list by excluding ingredients already available at home.

Grok Bot
Life & Hobbies

1–20 of 189

Work & ResearchWalkthrough

Summarize the Title and Description of New YouTube Videos Daily

Make fetches the title and description of new videos, summarizes them, and pastes the result into a Google Doc. It does not read the entire video or subtitles.

How it works

  1. Connect your Google account for use with YouTube and Google Docs in Make, and also register the connection information with the Hugging Face module. This setup requires permissions for video retrieval and document creation/editing.
  2. Configure the YouTube module to find the most recent video from a specific channel. The original example targeted @matthew_berman, and the values passed to the next step are Title and Description.
  3. Connect these two texts to the Hugging Face summarization module. While it is mentioned that you can adjust the summary length, the actual model used and the length setting are not disclosed.
  4. The first Google Docs module either creates a new document or finds an existing one to append content. The next module is set up to add the summary generated earlier to the end of the document.
  5. Set the scenario schedule to run every day at 9 AM. Each execution follows the sequence: fetch the latest video → summarize text → prepare document → append content.

Results

This is an example of automating the process of checking video information and transferring it to a document daily.

Notes

Only the title and description are summarized. The setup does not read the video content or subtitles. The original does not explain how to check for duplicates or set the time zone to avoid processing the same video again the next day.

Source

Labellerr 블로그 Original post ↗

Example prompt · Make

Set up a Make scenario that summarizes the title and description of new YouTube channel videos and accumulates them in Google Docs.

Input: Channel handle [channel], document to store [document], execution time and time zone [settings], summary length [condition]. First, check connections for YouTube, Hugging Face, and Google Docs, and investigate which summarization models and input lengths are actually available.

Fetch the latest video's title, description, URL, and identifier, and send only the title and description to the summarization model. Do not state that the entire video or subtitles were read. Add the result to the end of the document in the order: video title, link, summary, confirmation time.

Store the processed video identifier to prevent re-adding the same video the next day, and record the reason if the description is empty or a connection fails. Mark only items successfully saved to the document as completed.

Deliverables: Input/output connection table for each module, scenario file if possible, setup instructions. Test with new videos, repeated videos, empty descriptions, model failures, and document save failures, and compare the scheduled time zone and the actual content added to the document.