TimeVerse HyperFrames
FreeNot checkedWraps HyperFrames CLI as MCP tools, enabling AI agents to create, preview, and render video projects from HTML/CSS/JS, with text-to-speech and transcription cap
About
Wraps HyperFrames CLI as MCP tools, enabling AI agents to create, preview, and render video projects from HTML/CSS/JS, with text-to-speech and transcription capabilities.
README
将 HyperFrames CLI(HTML/CSS/JS 转视频渲染引擎)包装为 MCP 工具服务,让 AI Agent 可以直接:
- 创建/初始化视频项目
- 实时预览 HTML 视频(热重载)
- 渲染为 MP4/WebM
- 检查语法完整性
- 文本转语音 / 音频转录
- 检测运行环境、性能基准测试
架构概览
MCP 客户端(如 TimeVerseStudio)
│ (stdio 协议)
▼
┌─────────────────────────────────────┐
│ server.py (FastMCP) │
│ ┌─ 后台预热 npx 缓存(首次启动) │
│ └─ 注册 13 个 MCP 工具 │
└────────────┬────────────────────────┘
│
┌────────┼────────────┬──────────┐
▼ ▼ ▼ ▼
init.py render.py media.py system.py
lint.py preview.py
│
▼
┌─────────────────────────────┐
│ cli_executor.py │
│ ┌─ 优先全局 hyperframes │
│ │ 回退 npx hyperframes │
│ ├─ 超时/错误处理 │
│ └─ 环境检测 / lint 解析 │
└────────────┬────────────────┘
│
▼
hyperframes CLI (npx)
前置依赖
| 依赖 | 要求 | 安装 |
|---|---|---|
| Python | >= 3.11 | brew install [email protected] |
| Node.js | >= 22 | nodejs.org |
| FFmpeg | 任意版本 | macOS: brew install ffmpeg / Ubuntu: sudo apt install ffmpeg |
| HyperFrames | 自动通过 npx 获取 | 无需手动安装 |
首次运行时会自动通过 npx 下载 hyperframes 包(约 30-60 秒),后续缓存到本地。
可选优化:全局安装 hyperframes 可跳过 npx 下载步骤:
npm install -g hyperframes
安装
方式一:pip 安装(推荐)
pip install -e .
安装后可直接运行:
timeverse-hyperframes-mcp
方式二:uvx 运行(无需安装)
需要先安装 uv:
# macOS / Linux
curl -LsSf https://astral.sh/uv/install.sh | sh
# 直接通过 uvx 运行(自动下载依赖并缓存)
uvx timeverse-hyperframes-mcp
方式三:python -m
python -m timeverse_hyperframes_mcp.server
注册到 MCP 客户端
TimeVerseStudio
在 MCP 设置页面点击 JSON 导入,填入以下内容:
{
"mcpServers": {
"hyperframes": {
"description": "HyperFrames MCP — 将 HTML/CSS/JS 渲染为 MP4 视频。支持 init / preview / render / lint / TTS / transcribe",
"command": "uvx",
"args": ["timeverse-hyperframes-mcp"]
}
}
}
或通过 API 创建:
{
"name": "hyperframes",
"server_type": "local",
"transport": "stdio",
"command": "uvx",
"args": "timeverse-hyperframes-mcp",
"description": "HyperFrames MCP — 将 HTML/CSS/JS 渲染为 MP4 视频"
}
完整示例见 examples/timeverse-integration.json。
其他 MCP 客户端(如 Claude Desktop、Cline 等)
使用相同的 stdio 配置:
- 命令:
uvx timeverse-hyperframes-mcp或timeverse-hyperframes-mcp - 传输协议: stdio
可用工具(共 13 个)
项目初始化
| 工具 | 对应 CLI | 参数 | 说明 |
|---|---|---|---|
hyperframes_init |
npx hyperframes init |
project_name (必填), template, video_path, audio_path, parent_dir |
创建视频项目,支持 9 种模板 |
hyperframes_list_templates |
— | 无参数 | 列出所有模板及中文描述 |
支持模板:blank(空白)、warm-grain(暖色颗粒)、play-mode(Play 模式)、swiss-grid(瑞士网格)、vignelli(Vignelli 风格)、decision-tree(决策树)、kinetic-type(动态文字)、product-promo(产品宣传)、nyt-graph(纽约时报数据可视化)
视频渲染
| 工具 | 对应 CLI | 参数 | 说明 |
|---|---|---|---|
hyperframes_render |
npx hyperframes render |
output, fps(24/30/60), quality(draft/standard/high), format(mp4/webm), composition, workers(1-8/auto), gpu, docker, strict, strict_all |
渲染为 MP4/WebM,支持 GPU 加速和 Docker 可重现构建 |
webm格式支持透明背景- 渲染超时默认 10 分钟
实时预览
| 工具 | 对应 CLI | 参数 | 说明 |
|---|---|---|---|
hyperframes_preview |
npx hyperframes preview |
project_dir, port(默认 3002) |
启动本地预览服务,热重载,编辑 HTML 即时更新 |
语法检查
| 工具 | 对应 CLI | 参数 | 说明 |
|---|---|---|---|
hyperframes_lint |
npx hyperframes lint |
project_dir, verbose, json_output |
检查 data-composition-id、track 重叠、timeline 注册等问题 |
媒体处理
| 工具 | 对应 CLI | 参数 | 说明 |
|---|---|---|---|
hyperframes_tts |
npx hyperframes tts |
text (必填), voice, output, project_dir |
文本转语音,生成配音文件 |
hyperframes_list_voices |
npx hyperframes tts --list |
project_dir |
列出所有可用 TTS 语音角色(如 af_nova, bf_emma) |
hyperframes_transcribe |
npx hyperframes transcribe |
input_path (必填), model(tiny/base/small/medium/large), language, output_format, project_dir |
音频/视频转录为字幕 |
系统管理
| 工具 | 对应 CLI | 参数 | 说明 |
|---|---|---|---|
hyperframes_doctor |
npx hyperframes doctor |
project_dir |
检测 Node.js / npx / FFmpeg 环境,给出安装指引 |
hyperframes_info |
npx hyperframes info |
project_dir |
获取 HyperFrames 版本和环境详情 |
hyperframes_upgrade |
npx hyperframes upgrade |
check_only, json_output |
检查是否有新版本 |
hyperframes_compositions |
npx hyperframes compositions |
project_dir |
列出项目中的 composition 文件 |
hyperframes_benchmark |
npx hyperframes benchmark |
composition, project_dir |
运行渲染性能基准测试 |
典型工作流
flowchart LR
A[hyperframes_init] --> B[hyperframes_lint]
B --> C[hyperframes_preview]
C --> D[修改 HTML/CSS/JS]
D --> B
B --> E[hyperframes_render]
- 创建项目:
hyperframes_init(project_name="my-video", template="kinetic-type") - 语法检查:
hyperframes_lint(project_dir="my-video") - 实时预览:
hyperframes_preview(project_dir="my-video")→ 浏览器中编辑 HTML 即改即现 - 迭代: 修改 HTML → lint → 预览验证
- 渲染:
hyperframes_render(project_dir="my-video", quality="high")→ 输出 MP4
环境变量
| 变量 | 说明 | 默认值 |
|---|---|---|
HYPERFRAMES_WORKSPACE_DIR |
视频项目工作目录 | 当前工作目录 |
特性
- 智能命令解析:优先使用全局安装的
hyperframes,回退到npx hyperframes - 后台预热:服务器启动时自动在后台下载 hyperframes npm 缓存,首次调用更快
- 友好错误提示:命令超时或环境缺失时给出清晰的修复指引
- JSON 结构化输出:lint 结果支持 JSON 格式,便于 AI 解析
常见问题
Q: 首次运行很慢?
首次运行需要 npx 下载 hyperframes 包(约 30-60 秒),后续会缓存。也可以先全局安装 npm install -g hyperframes 跳过 npx 过程。
Q: 提示"未找到 npx"? 确认已安装 Node.js >= 22。
Q: 渲染失败?
先运行 hyperframes_doctor 检测环境是否完整,再运行 hyperframes_lint 检查语法。
开发者
- 源码:GitHub
- HyperFrames:https://github.com/heygen-com/hyperframes
Install TimeVerse HyperFrames in Claude Desktop, Claude Code & Cursor
unyly install timeverse-hyperframes-mcpInstalls into Claude Desktop, Claude Code, Cursor & VS Code — handles npx, uvx and build-from-source repos for you.
First time? Get the CLI: curl -fsSL https://unyly.org/install | sh
Or configure manually
Run in your terminal:
claude mcp add timeverse-hyperframes-mcp -- uvx timeverse-hyperframes-mcpFAQ
Is TimeVerse HyperFrames MCP free?
Yes, TimeVerse HyperFrames MCP is free — one-click install via Unyly at no cost.
Does TimeVerse HyperFrames need an API key?
No, TimeVerse HyperFrames runs without API keys or environment variables.
Is TimeVerse HyperFrames hosted or self-hosted?
Self-hosted: the server runs locally on your machine via the install command above.
How do I install TimeVerse HyperFrames in Claude Desktop, Claude Code or Cursor?
Open TimeVerse HyperFrames on unyly.org, pick your client tab (Claude Desktop, Claude Code, Cursor) and press Install — the config is generated automatically, no JSON editing.
Related MCPs
Omni Video
An MCP server that transforms LLM-enabled IDEs into professional video editors by pre-processing footage into text proxies, generating motion graphics via HTML/
by buildwithtazaARA
Generate images, video and audio from any AI agent — one connector.
by ARAYouTube
Transcripts, channel stats, search
by YouTubeEverArt
AI image generation using various models.
by modelcontextprotocolCompare TimeVerse HyperFrames with
Not sure what to pick?
Find your stack in 60 seconds
Author?
Embed badge for your README
Browse similar
All media MCPs
