More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    YouTube Transcript: Clean, Citable Text from Any Video

    1

    Turn any YouTube video into clean, usable text — fetch the best caption source, scrub the subtitle cruft, and deliver reading copies, timestamped indexes, or batch notes.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    YouTube Transcript: Clean, Citable Text from Any Video

    YouTube Transcript: Clean, Citable Text from Any Video

    Example session with this skill installed

    Get me clean, quotable text from this three-part lecture series, 40 minutes each, so I can cite the professor's key claim with its exact minute.

    • Read your context and instructions
    • Compiled the youtube transcript

    Reconnaissance first, never fetch blind: listed the available subtitle tracks and confirmed auto-generated English captions on all three lectures, plus the title and duration of each for file naming. Fetched all three with the best available source — uploader captions where they existed, auto captions where they did not — and kept the raw caption files in an archive folder next to the cleaned text, named after each video's title and ID. Cleaned each file: stripped timing metadata, HTML entities, duplicate lines from progressive captions, and music and applause annotations; merged single-word fragments into their neighbors; and ran a light sentence-splitting pass so the auto captions read as sentences instead of one long lowercase run. Delivered one notes file per lecture with a header (title, source, duration), the reading copy for human use, and a timestamped index built from the caption cue boundaries — one start-time plus text per unique cue — so quotes are citable to the minute. The key claim was found and cited at 00:23:41 in lecture two, with the surrounding lines kept for context. Total hands-on time: about five minutes for two hours of video, and the reading copies went straight into the summary step only after cleaning, since caption noise visibly degrades summary quality.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    YouTube is a library with no index. This skill gets the words out of a video and turns them into text you can actually work with — read, search, quote, summarize, or archive. Every job follows the same pipeline: fetch the best available caption source, clean the subtitle cruft out of it, and deliver the shape the purpose needs — an archive copy, a reading copy, or a timestamped index. Use it whenever a video's words need to become citable text, whether that is one lecture or a whole series at once.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Video Skills.

    It fetches the captions the uploader allows; a video with captions disabled goes to the re-recognition rung, which is slow and needs the user's confirmation first.
    Auto-translated captions are a rough gloss, never a quote source.
    Age-gated videos may need a signed-in session; private or deleted videos are gone, and the skill verifies the link instead of guessing.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 3 days ago

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Listed3 days ago

    What's inside

    Frequently Asked Questions

    Uploader-provided captions: human-written and punctuated. Then auto-generated captions, then auto-translated ones as a rough gloss. No captions at all means local re-recognition — the slow rung, confirmed with you first.

    Auto-caption noise visibly degrades summary quality. Cleaning first — duplicates, tag noise, sentence splits — means the summary works from real sentences, and the timestamps give it something to anchor claims to.

    The cleaned text with each cue's start time kept (00:03:12), built from the caption boundaries instead of throwing them away — so any line you quote carries its exact minute.

    New in Image & Video