Transcribe, repurpose and fact-check YouTube videos with Claude or ChatGPT connectors
A newsletter author shows three connector-based workflows: a video watcher that combines frames and transcript, a repurposer that creates posts and mini-courses, and a fact checker that verifies claims with web search.
Evidence: The author reports this. We have not checked it beyond reading the source.
The business problem
Creators and researchers want to reuse or verify long videos without watching them end to end or copying transcripts by hand.
What was tried
The author connects her own toolset to Claude or ChatGPT as a custom connector. The watcher downloads a video, extracts frames at intervals, pulls a timestamped transcript and combines them into one analysis (this part needs Cowork or Claude Code running locally). The repurposer reads only the transcript and produces a summary with quotes, posts for LinkedIn, Substack and X, a mini-course or an infographic. The fact checker takes a video, reel or claim, extracts factual claims, checks them with grounded web search and returns verdicts with links.
What was reported (mixed)
The author demonstrates a 13-minute video breakdown, a 2-hour podcast turned into a summary, posts, a mini-course and infographics, and a one-minute fact check that found most claims in a short video false. She notes the mini-course suited how-to content better than a discussion.
Limitations
The workflows are for premium subscribers of the author's newsletter, so this is partly a promotion. No accuracy figures are given for transcripts, summaries or verdicts, and the repurposer ignores visuals. Prompts must mention the connector or built-in tools are used instead, and OpenAI has not reviewed custom servers, so ChatGPT setup carries a risk the user accepts. Video length limits, privacy and plan requirements are not stated.
What you need
A Claude or ChatGPT account that supports custom connectors, a premium subscription to the author's newsletter, and Cowork or Claude Code for the frame-extraction workflow.
Sources
- AI Blew My Mind (Substack) ↗ Vendor case study, published June 21, 2026
Source published: June 21, 2026. Last reviewed here: October 11, 2026. Spot a mistake? Tell us.
Tools in this workflow
- Claude: Runs the connector workflowsTry Claude
- ChatGPT: Alternative host for the connectorsTry ChatGPT
Related workflows
Send a Monday business metrics email from Stripe, Notion and Sheets with Claude writing the narrative in n8n
A shared n8n template pulls last week's Stripe revenue, Notion deal pipeline and a Sheets row of support numbers, has Claude write a three-paragraph summary, and emails and posts it to Slack.
Set up Hermes Agent on a cheap VPS with Telegram, morning cron briefings and a reusable content research skill
A 14-day test of Hermes Agent on a small VPS, reached through Telegram, with memory files, a daily and weekly briefing, and custom skills for content research.
Chain Hermes agents through Telegram into a research, writing, editing and publishing team
A business-automation blog shows how to let several Hermes agents message each other in Telegram so a research agent can hand work to a writing agent in a content pipeline.
Draft help-centre-cited support replies and propose refunds that a person must approve
An open-source support workbench classifies tickets into three tiers, drafts replies that cite help-centre articles, reads billing data in read-only test mode, and only proposes refunds that need a named approver.