Flashcards

YouTube chapter markers extracted from the transcript by detecting topic shifts. Replaces manual scrubbing of the video for section boundaries.

A pause in a composing skill where it asks the user for input (e.g. pick a title) before continuing. Keeps the human in the loop on irreversible choices.

A one-click transcript generated by Clipchamp from a video file. The handoff point between video editing and the SharePoint AI skill pipeline.

The design principle that skills produce drafts and the human writes the final post. Avoids AI slop by keeping voice and judgment with the human.

A thin composing skill that detects keywords, runs the LinkedIn and YouTube skills, asks clarifying questions, and writes results back into SharePoint columns.

A skill trained on the author's prior posts that drafts a LinkedIn post matching his tone, format, and standard footer link block.

A reusable AI building block with a narrow job. Invoked on a file, returns structured output. Composes into workflows; trained on the specific format the user wants.

A pattern where one skill calls other skills as building blocks. Intelligence lives in the leaf skills; the composer handles routing and user input.

A technique of feeding several of the user's previous posts to Claude so it can extract tone, voice, and format — then encode that pattern into a skill.

A pattern where every downstream skill reads the video transcript rather than the raw video. Keeps skills cheap, grounded in spoken words, and easy to chain.

The pattern of saving skill output as column values on the SharePoint list item. Stages drafts for human review; nothing publishes automatically.

A skill that drafts a YouTube description from a transcript, suggests title candidates, and extracts chapter timestamps. The user picks the final title.