skills.rest
davepoon
davepoonCOMMUNITY
Explore all skills by this author →
Website𝕏@davepoonGitHub

video-to-text

Extract Douyin video metadata and transcribe speech from local video files.

3.4k|487|Updated Jul 25, 2025
Search Tags:#metadata extraction#FFmpeg#speech-to-text#faster-whisper#video transcription#Douyin#chinese transcription
One-click install
npx skills add https://github.com/davepoon/buildwithclaude --skill video-to-text
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: video-to-text
Source: https://github.com/davepoon/buildwithclaude/tree/main/plugins/all-skills/skills/video-to-text
Command: npx skills add https://github.com/davepoon/buildwithclaude --skill video-to-text

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires faster-whisper, and includes scripts (resource) components.

What problem does it solve?

Turning video content into searchable, copyable text is tedious: Douyin share links hide metadata behind login walls, and transcribing local videos requires manual audio extraction and speech recognition setup. This Skill automates both paths.

Core Features & Use Cases

  • Douyin Link Metadata Extraction: Parse Douyin share links via mobile SSR page parsing to get title, author, likes, hashtags, and duration without any Cookie or login.
  • Local Video Speech Transcription: Extract audio from local .mp4/.mov files using avconvert or ffmpeg, then transcribe Chinese speech with faster-whisper into timestamped text files.
  • Use Case: A user pastes a Douyin share link and asks to extract the video's copy and stats, or drops a local meeting recording and asks for a full Chinese transcript saved as a text file.

Quick Start

Use the video-to-text skill to extract the metadata from this Douyin share link, or transcribe the speech from my local video file into a timestamped text file.

Frequently Asked Questions about video-to-text

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract metadata from a Douyin share link without login?▼

Request the share link with a mobile iPhone User-Agent using curl, then parse the window._ROUTER_DATA JSON embedded in the SSR page. This returns title, author, likes, hashtags, and duration without any Cookie or login session.

How to transcribe speech from a local video file to text?▼

Extract the audio track with avconvert on macOS or ffmpeg on Linux, then run faster-whisper with the tiny model and language set to Chinese. The transcript is saved as a timestamped text file next to the input video.

Does Douyin link parsing require a Cookie or sessionid?▼

No, the primary SSR parsing method works without any Cookie by reading embedded page data. The older API endpoint is kept as a fallback and can use a DOUYIN_COOKIE environment variable if SSR parsing fails.

Why does Douyin metadata show duration of 0 seconds?▼

SSR metadata sometimes reports duration_ms as 0 even for real videos. If a download URL exists, download and verify the file first instead of assuming it is an image-text post; files under 1MB with zero duration are image posts without audio.

What are the limitations of faster-whisper video transcription?▼

Transcription is CPU-bound and slow: a 3-minute video takes about 5 minutes on an Intel Mac. Always set language to Chinese explicitly, since auto-detection adds 1-2 minutes, and the first run downloads a roughly 75MB tiny model.

Related Skills

Explore More Content & Communication Skills→
penghang1223penghang1223
1

douyin-video

Parse Douyin share links and download videos to a local workspace.

Community
Intermediate
levonklevonk

youtube-content-analysis

Extracts YouTube video transcripts and metadata for further processing.

Community
Intermediate
orthogonal-shorthogonal-sh
12

tiktok-search

Scrapes TikTok profiles, videos, and hashtags via Orthogonal's Scrape Creators API.

Official
Intermediate
HuaiminHuangHuaiminHuang

stt

Transcribe audio messages into text using a GPU-accelerated faster-whisper model.

Community
Intermediate