Adds a new community skill that combines video downloading (yt-dlp) with
speech-to-text transcription using fine-tuned Whisper models for 12 Indian
languages. The skill auto-routes to the best available model based on the
target language.
Supported languages: Telugu, Hindi, Kannada, Gujarati, Tamil, Bengali,
Malayalam, Marathi, Odia, Punjabi, Sanskrit, Urdu.
Models: vasista22/IIT Madras (HuggingFace) + AI4Bharat IndicWhisper.
Runs 100% locally — no API keys, no cloud services.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
a28aa032f7