video-to-subtitle-summary skill
A Codex and Claude Code skill: subtitles and a summary from a Douyin, Xiaohongshu, Bilibili or YouTube link or a local file
- 256 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The skill downloads video and writes files to disk
- Clip links go to the author's proxy service or TikHub, where your credits are spent
- Service keys sit in the `.env` file next to the skill
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/video-to-subtitle-summaryDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skill downloads video and writes files to disk
- Clip links go to the author's proxy service or TikHub, where your credits are spent
- Service keys sit in the `.env` file next to the skill
This is third-party code. Review the repository files before installing.
What it does
The skill turns a video into text with a summary. From a Douyin, Xiaohongshu, Bilibili or YouTube link it downloads the video or takes ready subtitles, and for a local file it first extracts the audio. It then builds subtitles, asks the agent for a short summary and key points and saves the results in a folder: video, audio, SRT subtitles and plain text. By default speech recognition runs locally through `faster-whisper`, with no cloud sign-up. The variable `ASR_BACKEND=volcengine` switches to the Volcengine subtitle service. For YouTube the skill first tries to download author or automatic subtitles through `yt-dlp` and skips recognition. Douyin, Xiaohongshu and Bilibili clips need a direct file link, which comes from the author's paid "AI Douyin" proxy or your own TikHub account.
Who it is for. Content managers and researchers who go through short videos from Chinese social networks and YouTube and want text and a summary.
Good fit when
- You have a YouTube link and want its subtitles and a short summary without running recognition
- You need to transcribe a local call or lecture recording into SRT and plain text
- You must go through a dozen short Douyin clips quickly and get the key points of each
- You want fully local transcription with no cloud services or recognition fees
Not a fit when
- You want to download someone else's video to republish it: the skill prepares text for analysis and the rights stay with the author
- You do not want to pay for a proxy: Douyin, Xiaohongshu and Bilibili then need your own TikHub account
- You need a Russian translation: the agent writes the summary, usually in the language of the request
Example request
Pull the subtitles from this YouTube video, write a short summary with key points and give me the subtitle file path separatelyLimitations
Python 3.9 or newer, `ffmpeg` and `yt-dlp` for Bilibili and YouTube are required. For Douyin, Xiaohongshu and Bilibili the direct file link comes from the author's "AI Douyin" service, where a successful parse costs one credit after a free allowance, or from your own TikHub token. Keys go in `.env`. Volcengine cloud recognition needs a separate account and keys. Documentation is mostly in Chinese, with a short English version. Downloading other people's videos is subject to platform rules and copyright.
How to disable. Delete the `video-to-subtitle-summary` folder from `~/.codex/skills` or `~/.claude/skills` and the `.env` file next to it. Downloaded videos and subtitles sit in the output folder, which you can clear by hand.
FAQ
How do I install the skill?
Clone the repository and copy the folder to `~/.codex/skills` or `~/.claude/skills` under the name `video-to-subtitle-summary`, then copy `.env.example` to `.env`. The author also suggests asking an agent to install it from the repository link.
Do I need a paid recognition service?
No, local `faster-whisper` is the default. Volcengine is enabled with `ASR_BACKEND=volcengine` and needs a token and app ID.
What does the skill do with YouTube?
It first tries to download author or automatic subtitles through `yt-dlp`. The video stays on YouTube and recognition is skipped.
How does the skill fetch Douyin and Xiaohongshu clips?
Through the author's "AI Douyin" service, where new users get a free allowance, or through your TikHub token in the `TIKHUB_TOKEN` variable.
Which files come out?
Video, mp3 audio, SRT subtitles and plain text, plus a summary and key points in the agent's reply.
Discussion
No questions or reviews yet.
Related
SkillsA skill that turns a Bilibili, Douyin or YouTube video link into timestamped notes through local transcription
A Claude Code skill that downloads a YouTube video, splits it into semantic chapters, cuts chosen clips, translates subtitles and burns bilingual ones in
A skill that transcribes video, lets you review captions in a browser and burns in Chinese and English captions.
A skill that turns a script into a finished video with voice-over, subtitles, a cover and a publish text, frames drawn by code.
An agent edits short talking-head videos: transcript, subtitles, B-roll and export for Xiaohongshu, Douyin and WeChat Channels
Skill that turns a video into tall subtitle collages: crops burned-in subtitles or draws reviewed lines onto real frames
README in short
The project description shows two routes: a clip link or a local file. A one-line install through an agent, the choice of recognition engine, a dependency table, installation steps and a sample result follow. At the end the author offers an online version and private deployment for companies.