mirror of
https://github.com/hanzoskill/gettr-transcribe-summarize.git
synced 2026-10-30 21:10:53 +00:00
853 B
853 B
Notes / Reference
This skill assumes a simple, mostly-static GETTR post HTML page where the media URL is discoverable via og:video* meta tags.
Meta tags searched
The extractor script prefers these, in order:
og:video:secure_urlog:video:urlog:video
Output directory
./out/gettr-transcribe-summarize/<slug>/
The slug is extracted from the GETTR post URL path (e.g., /post/p1abc2def → p1abc2def).
Output files
audio.wav– 16kHz mono audio (optimized for Whisper)audio.vtt– timestamped transcript (VTT format)summary.md– final deliverable with bullets and optional timestamped outline
Exit codes (extract script)
0: success (video URL and slug printed to stdout)1: no video found (post may be text/image only)2: usage error or invalid URL3: network error after retries