Install
mkdir -p .claude/skills/videocut && curl -L -o skill.zip "https://mcp.directory/api/skills/download/5017" && unzip -o skill.zip -d .claude/skills/videocut && rm skill.zipInstalls to .claude/skills/videocut
About this skill
剪口播 v2
火山引擎转录 + AI 口误识别 + 网页审核
快速使用
用户: 帮我剪这个口播视频
用户: 处理一下这个视频
输出目录结构
output/
└── YYYY-MM-DD_视频名/
├── 剪口播/
│ ├── 1_转录/
│ │ ├── audio.mp3
│ │ ├── volcengine_result.json
│ │ └── subtitles_words.json
│ ├── 2_分析/
│ │ ├── readable.txt
│ │ ├── auto_selected.json
│ │ └── 口误分析.md
│ └── 3_审核/
│ └── review.html
└── 字幕/
└── ...
规则:已有文件夹则复用,否则新建。
流程
0. 创建输出目录
↓
1. 提取音频 (ffmpeg)
↓
2. 上传获取公网 URL (uguu.se)
↓
3. 火山引擎 API 转录
↓
4. 生成字级别字幕 (subtitles_words.json)
↓
5. AI 分析口误/静音,生成预选列表 (auto_selected.json)
↓
6. 生成审核网页 (review.html)
↓
7. 启动审核服务器,用户网页确认
↓
【等待用户确认】→ 网页点击「执行剪辑」或手动 /剪辑
执行步骤
步骤 0: 创建输出目录
# 变量设置(根据实际视频调整)
VIDEO_PATH="/path/to/视频.mp4"
VIDEO_NAME=$(basename "$VIDEO_PATH" .mp4)
DATE=$(date +%Y-%m-%d)
BASE_DIR="output/${DATE}_${VIDEO_NAME}/剪口播"
# 创建子目录
mkdir -p "$BASE_DIR/1_转录" "$BASE_DIR/2_分析" "$BASE_DIR/3_审核"
cd "$BASE_DIR"
步骤 1-3: 转录
cd 1_转录
# 1. 提取音频(文件名有冒号需加 file: 前缀)
ffmpeg -i "file:$VIDEO_PATH" -vn -acodec libmp3lame -y audio.mp3
# 2. 上传获取公网 URL
curl -s -F "files[]=@audio.mp3" https://uguu.se/upload
# 返回: {"success":true,"files":[{"url":"https://h.uguu.se/xxx.mp3"}]}
# 3. 调用火山引擎 API
SKILL_DIR="/Users/chengfeng/Desktop/AIos/剪辑Agent/.claude/skills/剪口播"
"$SKILL_DIR/scripts/volcengine_transcribe.sh" "https://h.uguu.se/xxx.mp3"
# 输出: volcengine_result.json
步骤 4: 生成字幕
node "$SKILL_DIR/scripts/generate_subtitles.js" volcengine_result.json
# 输出: subtitles_words.json
cd ..
步骤 5: 分析口误(脚本+AI)
5.1 生成易读格式
cd 2_分析
node -e "
const data = require('../1_转录/subtitles_words.json');
let output = [];
data.forEach((w, i) => {
if (w.isGap) {
const dur = (w.end - w.start).toFixed(2);
if (dur >= 0.5) output.push(i + '|[静' + dur + 's]|' + w.start.toFixed(2) + '-' + w.end.toFixed(2));
} else {
output.push(i + '|' + w.text + '|' + w.start.toFixed(2) + '-' + w.end.toFixed(2));
}
});
require('fs').writeFileSync('readable.txt', output.join('\\n'));
"
5.2 读取用户习惯
先读 用户习惯/ 目录下所有规则文件。
5.3 生成句子列表(关键步骤)
必须先分句,再分析。按静音切分成句子列表:
node -e "
const data = require('../1_转录/subtitles_words.json');
let sentences = [];
let curr = { text: '', startIdx: -1, endIdx: -1 };
data.forEach((w, i) => {
const isLongGap = w.isGap && (w.end - w.start) >= 0.5;
if (isLongGap) {
if (curr.text.length > 0) sentences.push({...curr});
curr = { text: '', startIdx: -1, endIdx: -1 };
} else if (!w.isGap) {
if (curr.startIdx === -1) curr.startIdx = i;
curr.text += w.text;
curr.endIdx = i;
}
});
if (curr.text.length > 0) sentences.push(curr);
sentences.forEach((s, i) => {
console.log(i + '|' + s.startIdx + '-' + s.endIdx + '|' + s.text);
});
" > sentences.txt
5.4 脚本自动标记静音(必须先执行)
node -e "
const words = require('../1_转录/subtitles_words.json');
const selected = [];
words.forEach((w, i) => {
if (w.isGap && (w.end - w.start) >= 0.5) selected.push(i);
});
require('fs').writeFileSync('auto_selected.json', JSON.stringify(selected, null, 2));
console.log('≥0.5s静音数量:', selected.length);
"
→ 输出 auto_selected.json(只含静音 idx)
5.5 AI 分析口误(追加到 auto_selected.json)
检测规则(按优先级):
| # | 类型 | 判断方法 | 删除范围 |
|---|---|---|---|
| 1 | 重复句 | 相邻句子开头≥5字相同 | 较短的整句 |
| 2 | 隔一句重复 | 中间是残句时,比对前后句 | 前句+残句 |
| 3 | 残句 | 话说一半+静音 | 整个残句 |
| 4 | 句内重复 | A+中间+A 模式 | 前面部分 |
| 5 | 卡顿词 | 那个那个、就是就是 | 前面部分 |
| 6 | 重说纠正 | 部分重复/否定纠正 | 前面部分 |
| 7 | 语气词 | 嗯、啊、那个 | 标记但不自动删 |
核心原则:
- 先分句,再比对:用 sentences.txt 比对相邻句子
- 整句删除:残句、重复句都要删整句,不只是删异常的几个字
分段分析(循环执行):
1. Read readable.txt offset=N limit=300
2. 结合 sentences.txt 分析这300行
3. 追加口误 idx 到 auto_selected.json
4. 记录到 口误分析.md
5. N += 300,回到步骤1
🚨 关键警告:行号 ≠ idx
readable.txt 格式: idx|内容|时间
↑ 用这个值
行号1500 → "1568|[静1.02s]|..." ← idx是1568,不是1500!
口误分析.md 格式:
## 第N段 (行号范围)
| idx | 时间 | 类型 | 内容 | 处理 |
|-----|------|------|------|------|
| 65-75 | 15.80-17.66 | 重复句 | "这是我剪出来的一个案例" | 删 |
步骤 6-7: 审核
cd ../3_审核
# 6. 生成审核网页
node "$SKILL_DIR/scripts/generate_review.js" ../1_转录/subtitles_words.json ../2_分析/auto_selected.json ../1_转录/audio.mp3
# 输出: review.html
# 7. 启动审核服务器
node "$SKILL_DIR/scripts/review_server.js" 8899 "$VIDEO_PATH"
# 打开 http://localhost:8899
用户在网页中:
- 播放视频片段确认
- 勾选/取消删除项
- 点击「执行剪辑」
数据格式
subtitles_words.json
[
{"text": "大", "start": 0.12, "end": 0.2, "isGap": false},
{"text": "", "start": 6.78, "end": 7.48, "isGap": true}
]
auto_selected.json
[72, 85, 120] // Claude 分析生成的预选索引
配置
火山引擎 API Key
cd /Users/chengfeng/Desktop/AIos/剪辑Agent/.claude/skills
cp .env.example .env
# 编辑 .env 填入 VOLCENGINE_API_KEY=xxx
You might also like
flutter-development
aj-geddes
Build beautiful cross-platform mobile apps with Flutter and Dart. Covers widgets, state management with Provider/BLoC, navigation, API integration, and material design.
drawio-diagrams-enhanced
jgtolentino
Create professional draw.io (diagrams.net) diagrams in XML format (.drawio files) with integrated PMP/PMBOK methodologies, extensive visual asset libraries, and industry-standard professional templates. Use this skill when users ask to create flowcharts, swimlane diagrams, cross-functional flowcharts, org charts, network diagrams, UML diagrams, BPMN, project management diagrams (WBS, Gantt, PERT, RACI), risk matrices, stakeholder maps, or any other visual diagram in draw.io format. This skill includes access to custom shape libraries for icons, clipart, and professional symbols.
ui-ux-pro-max
nextlevelbuilder
"UI/UX design intelligence. 50 styles, 21 palettes, 50 font pairings, 20 charts, 8 stacks (React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind). Actions: plan, build, create, design, implement, review, fix, improve, optimize, enhance, refactor, check UI/UX code. Projects: website, landing page, dashboard, admin panel, e-commerce, SaaS, portfolio, blog, mobile app, .html, .tsx, .vue, .svelte. Elements: button, modal, navbar, sidebar, card, table, form, chart. Styles: glassmorphism, claymorphism, minimalism, brutalism, neumorphism, bento grid, dark mode, responsive, skeuomorphism, flat design. Topics: color palette, accessibility, animation, layout, typography, font pairing, spacing, hover, shadow, gradient."
godot
bfollington
This skill should be used when working on Godot Engine projects. It provides specialized knowledge of Godot's file formats (.gd, .tscn, .tres), architecture patterns (component-based, signal-driven, resource-based), common pitfalls, validation tools, code templates, and CLI workflows. The `godot` command is available for running the game, validating scripts, importing resources, and exporting builds. Use this skill for tasks involving Godot game development, debugging scene/resource files, implementing game systems, or creating new Godot components.
nano-banana-pro
garg-aayush
Generate and edit images using Google's Nano Banana Pro (Gemini 3 Pro Image) API. Use when the user asks to generate, create, edit, modify, change, alter, or update images. Also use when user references an existing image file and asks to modify it in any way (e.g., "modify this image", "change the background", "replace X with Y"). Supports both text-to-image generation and image-to-image editing with configurable resolution (1K default, 2K, or 4K for high resolution). DO NOT read the image file first - use this skill directly with the --input-image parameter.
fastapi-templates
wshobson
Create production-ready FastAPI projects with async patterns, dependency injection, and comprehensive error handling. Use when building new FastAPI applications or setting up backend API projects.
Stay ahead of the MCP ecosystem
Get weekly updates on new skills and servers.