Transcribe and Summarize Content From Multiple Formats

Extracts and summarizes content from diverse sources so you can understand key points without reading entire documents or watching full videos. It handles YouTube transcripts, PDFs, Word docs, spreadsheets, web articles, and audio files, then organizes the output with key points, quotes, and trends. This removes the friction of switching between tools to process different formats.

Once installed, you provide a content source (URL, file path, or video link) and describe what you need. The skill automatically detects the format, extracts content using the appropriate method, and delivers a structured summary with key points, important quotes, relevant data trends, and timestamps where applicable. For comparison workflows, it summarizes multiple sources and creates side-by-side analyses.

Key points from any content format surface in one place with consistent structure, so you skip the manual work of converting, reading, and extracting insights across formats.

Requires Python 3.6+ for CLI use. YouTube transcript extraction requires youtube-transcript-api (installed via pip) to avoid manual fallback. PDF and document parsing uses Claude's built-in Read and WebFetch tools. Works with Claude Web (no setup) or Claude Code (local file access). The skill includes helper scripts at scripts/youtube_transcript.py for standalone YouTube extraction.

YouTube transcription requires internet access and either youtube-transcript-api library or a working helper script. Scanned PDFs (image-based) require external OCR preprocessing. Paywalled content and restricted articles cannot be accessed. Large files may require section-by-section processing. Direct audio transcription requires pre-existing transcript files or external transcription services (Whisper, Rev, etc.).