A robust transcript extraction utility that uses watch-page scraping and InnerTube API fallbacks to capture captions when other tools fail.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install youtube-transcript-fetcher
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install youtube-transcript-fetcher using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is YouTube Transcript Fetcher?
The YouTube Transcript Fetcher is a specialized utility designed for developers and researchers who need reliable access to video text data. Unlike standard caption tools that often fail when uploader subtitles are missing, this skill utilizes a multi-layered approach to ensure high recovery rates for auto-generated captions. By integrating this into your Openclaw Skills library, you gain a powerful asset for data collection and content analysis.
At its core, the tool prioritizes raw transcript accuracy over summarization. It is built to handle complex scenarios by rotating through multiple InnerTube player API client profiles, including Android, Web, and TV, to find valid caption tracks that simpler methods might miss.
YouTube Transcript Fetcher Use Cases
- Extracting source transcript text for RAG (Retrieval-Augmented Generation) and AI training data.
- Batch processing recent video transcripts from specific YouTube channels for market research.
- Recovering hard-to-reach auto-generated captions for accessibility and indexing.
- Automating daily content monitoring through structured configuration files.
How YouTube Transcript Fetcher Works
- The user provides a single YouTube URL, a channel handle, or a batch configuration file.
- The script first attempts to scrape the transcript metadata directly from the YouTube watch page.
- If initial scraping fails, the system triggers a fallback mechanism using the InnerTube player API.
- It iterates through multiple client profiles (ANDROID, WEB, TVHTML5_SIMPLY_EMBEDDED_PLAYER, IOS) to locate available caption tracks.
- The tool processes the raw caption data and returns a structured JSON output containing the full transcript text.
YouTube Transcript Fetcher Setup
To integrate this with your existing Openclaw Skills, ensure you have the required system dependencies installed. First, install the yt-dlp binary:
# For macOS users
brew install yt-dlp
# For Debian/Ubuntu users
sudo apt install yt-dlp
Then, navigate to the skill directory and install the necessary Python dependencies:
pip install -r requirements.txt
You can then run the fetcher using:
./youtube-transcript-fetcher --url <youtube-url>
YouTube Transcript Fetcher Data Schema & Taxonomy
The skill generates structured data focusing on the raw text content. The following schema describes the typical output structure:
| Attribute | Description |
|---|---|
transcript |
The full raw text recovered from the video captions. |
video_id |
The unique 11-character identifier for the video. |
method |
The successful extraction method (e.g., scraping or specific InnerTube profile). |
status |
Confirmation of successful retrieval or specific error code. |
YouTube Transcript Fetcher Advanced Features
- Multi-client InnerTube fallback system to bypass regional or platform-specific caption restrictions.
- Batch processing mode using
channels.jsonfor large-scale data ingestion. - Time-aware channel crawling using the
--hoursflag to fetch only recent content. - Headless operation capability designed for seamless integration within Openclaw Skills automated workflows.
SKILL.md
Loading
Related Openclaw Skills
Text Tighten
A high-performance text optimization tool designed to sharpen writing, remove fluff, and improve structural clarity without losing original intent.
Seedance 2.0 Storyboard Director
A professional director skill for crafting movie-grade AI video prompts and cinematic storyboards from multimodal inputs.
OpenClaw Permissions Audit Skill
A security-focused utility to audit, review, and report on the permissions, API access, and communication channels utilized by an AI agent.
Video Script Writer Claw
A professional AI scriptwriter designed to generate shot-by-shot video scripts, podcast outlines, and multi-episode content plans with brand-consistent styling.
Image Creator
A sophisticated image generation assistant specializing in 17 curated 3D styles across characters, scenes, and commercial product designs.
incident-response-team-setup
A comprehensive 7-phase framework for building, documenting, and validating an incident response team and its operational testing program.
Featured*
