rip_transcript keeps the full text out of the response unless return_text is set, on the stated reasoning that it keeps context lean and leaves a persistent file that can be reused. What comes back instead is enough to decide whether to open it: title, channel, language, duration, word count and the saved path. The vendor also asks that the saved location always be reported to the user afterwards, which is a fair habit to keep when the output lands outside the conversation.
A single-tool YouTube transcript extractor that saves the transcript as a Markdown file and returns a resource link with metadata rather than the text itself.
- rip_transcript extracts a video's full transcript and, by default, writes it to a Markdown file, returning a resource link plus the title, channel, language, duration, word count, saved path and a preview.
- return_text overrides that default when the transcript itself is what you want back, and format and save_path control the output form and where it lands.
A writable location for the saved file — save_path sets it, and the tool reports the path it used.
One command — npx rippr-mcp