DEDUPED Hot AI repos, turned into use cases Get the Monday 5

How to find the exact second viewers leave a video with ChatGPT YouTube Agent Skill

Read a retention export and name the second people left and what was said.

  • Content
  • Intermediate
  • 15 minutes
  • From the README
  • Codex

Who it is forA YouTube editor reviewing a published video whose audience keeps dropping off early.

The problemA retention graph shows that viewers left, but not which moment in the script caused it.

The outcomeA list of drop-off points with the exact second and the matching line of speech, so you know what to fix in the next edit.

What you need first

  • ChatGPT or the Codex app that loads skills from ~/.codex/skills/
  • git installed, if you use the clone option
  • Python 3 to run the bundled scripts (the README says they need no packages)
  • Your own channel data, such as an audience-retention export, for the analysis skills

Steps

  1. Install the skills using the link or the git clone and copy commands in the README.
  2. Restart the app.
  3. Export the audience-retention data for the video from your channel.
  4. Ask the yt-retention skill to read the export and name the seconds people left.
  5. Run python retention.py on the export file for a local check of the drop points. (inferred, check the README)

Prompt for your agent

Using the chatgpt-youtube-agent-skill repo, run the yt-retention skill on the retention export at <export-file> for the video <video-title>. Expected output: a list of exact seconds where viewers left, each with the spoken words at that point. Verify that every listed second matches a timestamp present in the export.

Caveat. The README describes the retention skill as reading your export, so its accuracy depends on the export containing per-second data and the matching transcript.

Share: X LinkedIn Hacker News Email Copy link

Disclaimer

Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked "Seen in the wild", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email [email protected] and we will look at it promptly. Full disclaimer page