artemchuikin/youtube-skills
YouTube transcript skills for Claude Code, Cursor and Codex. Agent Skills to get transcripts, search YouTube, read channels and playlists, and bulk export.
What it solves
This project provides a set of pre-configured "skills" that allow AI coding and chat agents to interact with YouTube directly. It eliminates the need for complex workarounds like headless browsers or managing Google Cloud projects, enabling agents to fetch transcripts, search videos, and browse channels or playlists with simple natural language prompts.
How it works
These skills are folders of instructions that your AI agent loads to understand how to communicate with the TranscriptOut API. When you ask your agent a question—such as "What does this video say?"—the agent uses the loaded skill to format the request, handle errors, and process the data. The setup process is automated: the agent can guide you through creating a free account and saving your API key directly within your chat session.
Who it’s for
It is designed for users of AI agents like Claude Code, Cursor, Cline, and others who want to integrate YouTube data into their workflows without manual scripting or managing API quotas.
Highlights
- Comprehensive Toolkit: Includes 12 specialized skills for tasks like transcript extraction, channel browsing, playlist management, and bulk data exports.
- Bulk Processing: Supports efficient batch jobs for large-scale tasks, such as archiving entire channels, without needing to write custom loops.
- Retrieval-Ready: Transcripts can be returned in pre-chunked segments with precise, word-level timestamps, making them ready for use in RAG (Retrieval-Augmented Generation) pipelines.
- No Google API Required: Operates independently of Google’s YouTube Data API, avoiding quota issues and complex authentication requirements.
- Easy Setup: Offers a seamless, in-chat installation flow that handles API key generation and configuration automatically.
Related
- Project
- Project
- Project
- Project