Videocut Subtitle
- 481 installs
- 263 repo stars
- Updated July 22, 2026
- zrt-ai-lab/opencode-skills
videocut-subtitle is an agent skill that automates Whisper transcription, dictionary correction, SRT generation, and FFmpeg subtitle burn-in for developers who need captioned MP4 files without hand-aligning timestamps.
About
videocut-subtitle is an agent skill from zrt-ai-lab/opencode-skills videocut series (version 1.0.0) that runs the full Chinese subtitle pipeline: Whisper speech-to-text, dictionary-based error correction, human review, SRT export, and FFmpeg subtitle burning. It sits in the videocut family alongside install, clip-oral, clip, and self-update skills that share ffmpeg and Python 3.10+ dependencies from the parent repository. Each run produces three artifacts—a plain-text 字幕稿 for editing, a timed .srt file, and a burned-in 字幕.mp4. Agents trigger on phrases like 加字幕 or 生成字幕, typically after videocut-clip finishes so timestamps match the edited cut. Developers reach for videocut-subtitle inside OpenCode or Claude Code when batch-producing training, marketing, or tutorial videos and want dictionary hotwords, review checkpoints, and FFmpeg force_style captions without writing filter graphs manually.
- Full pipeline: transcription → correction → review → SRT → burn
- Uses Whisper (medium/large-v3) for speech-to-text
- Dictionary-based correction for proper nouns and brand names
- FFmpeg integration for final subtitle burning
- Strict formatting: one line per screen, max 15 chars
Videocut Subtitle by the numbers
- 481 all-time installs (skills.sh)
- Ranked #414 of 2,719 Automation & Workflows skills by installs in the Skillselion catalog
- Data as of Aug 3, 2026 (Skillselion catalog sync)
npx skills add https://github.com/zrt-ai-lab/opencode-skills --skill videocut-subtitleAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 481 |
|---|---|
| repo stars | ★ 263 |
| Last updated | July 22, 2026 |
| Repository | zrt-ai-lab/opencode-skills ↗ |
How do you automate Whisper subtitles and FFmpeg burn-in?
Automates the full subtitle pipeline for video: Whisper transcription, dictionary-based error correction, user review, SRT generation, and FFmpeg subtitle burning.
Who is it for?
Developers producing captioned Chinese video inside agent sessions who already use the zrt-ai-lab videocut clip workflow and need Whisper plus FFmpeg wired together.
Skip if: Teams needing real-time live captions, non-Chinese-only pipelines without dictionary correction, or video editing without ffmpeg and Python 3.10+ installed.
When should I use this skill?
User asks to add subtitles, generate SRT from video, run Whisper transcription, burn captions with FFmpeg, or mentions videocut-subtitle after clipping.
What you get
Plain-text 字幕稿, timed .srt file, and burned-in 字幕.mp4 with dictionary-corrected captions.
- 字幕稿 text draft
- SRT subtitle file
- Burned-in captioned MP4
By the numbers
- Ships as version 1.0.0 in the zrt-ai-lab videocut skill family
- Produces 3 output artifacts: 字幕稿.txt, .srt, and 字幕.mp4
- Parent videocut series documents 6 related videocut skills sharing ffmpeg
Files
字幕
转录 → 纠错 → 审核 → 匹配 → 烧录
流程
1. 转录视频（Whisper）
↓
2. 词典纠错 + 分句
↓
3. 输出字幕稿（纯文本，一句一行）
↓
【用户审核修改】
↓
4. 用户给回修改后的文本
↓
5. 我匹配时间戳 → 生成 SRT
↓
6. 烧录字幕（FFmpeg）转录
使用 OpenAI Whisper 模型进行语音转文字：
whisper video.mp4 --model medium --language zh --output_format json| 模型 | 用途 |
|---|---|
medium | 默认，平衡速度与准确率 |
large-v3 | 高精度，较慢 |
输出 JSON 包含逐词时间戳，用于后续 SRT 生成。
---
字幕规范
| 规则 | 说明 |
|---|---|
| 一屏一行 | 不换行，不堆叠 |
| ≤15字/行 | 超过15字必须拆分（4:3竖屏） |
| 句尾无标点 | 你好 不是 你好。 |
| 句中保留标点 | 先点这里，再点那里 |
---
词典纠错
读取 词典.txt，每行一个正确写法：
skills
Claude
iPhone我自动识别变体：claude → Claude
---
字幕稿格式
我给用户的（纯文本，≤15字/行）：
今天给大家分享一个技巧
很多人可能不知道
其实这个功能
藏在设置里面
你只要点击这里
就能看到了用户修改后给回我，我再匹配时间戳生成 SRT。
---
样式
默认：24号白字、黑色描边、底部居中
可选样式：
| 样式 | 说明 |
|---|---|
| 默认 | 白字黑边 |
| 黄字 | 黄字黑边（醒目） |
用户可说：
- "字大一点" → 32号
- "放顶部" → 顶部居中
- "黄色字幕" → 黄字黑边
---
输出
01-xxx_字幕稿.txt # 纯文本，用户编辑
01-xxx.srt # 字幕文件
01-xxx-字幕.mp4 # 带字幕视频videocut:字幕
字幕生成与烧录
文件
| 文件 | 作用 |
|---|---|
SKILL.md | 流程定义 |
词典.txt | 正确写法列表（每行一个） |
词典格式
skills
Claude
钉钉AI录音卡用户只写正确的词，我识别所有错误变体。
流程
转录 → 词典纠错 → 用户审核 → 烧录skills
Claude
iPhone
GitHub
API
APP
NO.1
钉钉
Related skills
Agent BrowserGive their coding agent reliable, high-fidelity control over any website or Electron desktop app.591k39.6k
Lark ApprovalLet their AI coding agent create, read, update, and approve items in Lark (Feishu) approval workflows without leaving the coding environment.494kLark EventHandle Feishu/Lark bot events, webhooks, and subscription callbacks in agent-driven backend code.493kLark Workflow Meeting SummaryAutomatically generate structured meeting summaries and action items from Lark/Feishu calls without manual note-taking.492kLark Workflow Standup ReportAutomatically generate and post daily standup reports in Lark/Feishu from their workflow and activity data.492k
Twitter AutomationAutomate X/Twitter posts, scheduling, and engagement via X API integration441k669How it compares
Pick videocut-subtitle for full Whisper-to-burned-caption pipelines; use video-subtitle-remover in the same repo when hard subtitles or watermarks must be stripped instead.
FAQ
What tools does videocut-subtitle orchestrate?
videocut-subtitle orchestrates OpenAI Whisper for transcription, a dictionary file for hotword correction, human review of the 字幕稿, SRT generation, and FFmpeg subtitles filters to burn captions into MP4.
What files does videocut-subtitle produce?
videocut-subtitle outputs a plain-text 字幕稿 for editing, a timed .srt subtitle file, and a final 字幕.mp4 video with burned-in captions after review and FFmpeg processing.
When should videocut-subtitle run in the videocut workflow?
videocut-subtitle should run after videocut-clip finishes, because clip edits remove segments and original Whisper timestamps will not align with the cut video.