From vtt-to-text
This skill should be used when the user pastes WebVTT (.vtt) closed caption content and wants it converted into a clean text file. Trigger when the user pastes content that starts with "WEBVTT" or contains VTT timestamp arrow patterns between timecodes, asks to "clean up captions", "convert subtitles to text", "extract text from VTT", "turn captions into a transcript", or wants to turn video closed captions into readable text. Also trigger when the user mentions converting course lecture captions or subtitle files.
How this skill is triggered — by the user, by Claude, or both
Slash command
/vtt-to-text:vtt-to-textThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Convert pasted WebVTT closed caption content into clean, readable plain text files.
Convert pasted WebVTT closed caption content into clean, readable plain text files.
The user will paste VTT content directly into the conversation. Recognizable by:
WEBVTT header line--> arrows (e.g., 00:00:03.885 --> 00:00:05.395)Before processing, ask the user for three things using AskUserQuestion:
The final filename pattern is: {Tier} - {Sprint} - {Title}.txt
Example: Tier 3 - Sprint 1 - Introduction.txt
If the user has already provided any of these details in their message, skip those questions and use what they gave.
Save the raw VTT content to a temp file and run the bundled parser:
python3 ${CLAUDE_PLUGIN_ROOT}/skills/vtt-to-text/scripts/parse_vtt.py < /tmp/raw_vtt.txt > /tmp/cleaned.txt
The parser:
WEBVTT headerRead the cleaned output, save it to the user's workspace folder with the filename from Step 2, and provide a computer:// link.
Keep the response brief — confirm the file was created and link it. No need to show the full text content.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub moxywolfllc/moxywolf-plugins --plugin vtt-to-text