Figma Context MCP
Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.
作者提供的说明(英文原文)
Transcribe audio/video files or URLs via the WhipScribe cloud API and fetch the transcript.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
WHIPSCRIBE_API_KEY标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"whipscribe": {
"command": "uvx",
"args": [
"whipscribe-mcp"
],
"env": {
"WHIPSCRIBE_API_KEY": "your-value"
}
}
}
}在终端中运行这条命令
claude mcp add whipscribe --env WHIPSCRIBE_API_KEY=your-value -- uvx whipscribe-mcp运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"whipscribe": {
"command": "uvx",
"args": [
"whipscribe-mcp"
],
"env": {
"WHIPSCRIBE_API_KEY": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"whipscribe": {
"type": "stdio",
"command": "uvx",
"args": [
"whipscribe-mcp"
],
"env": {
"WHIPSCRIBE_API_KEY": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月8日
注册名称已通过 GitHub 账户 neugence 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
你需要填写:WHIPSCRIBE_API_KEY
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(0.1.5)于35天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Python (uv)。
同一分类中有:Figma Context MCP, Claude Real Video, Figma MCP Server
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Give your coding agent access to your Figma data. Implement designs in any framework in one-shot.
Let any LLM watch a video locally — and search everything it has ever watched.
The Figma MCP server brings Figma design context directly into your AI workflow.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
mcp-name: io.github.neugence/whipscribe-mcp
MCP server for Whipscribe — transcribe audio and video from a URL or local file via Claude Desktop, Claude Code, Cursor, Windsurf, or any MCP-compatible client.
### ⚠️ Beta service
-
whipscribe-mcpand the Whipscribe API are in beta. Endpoints, response shapes, and quotas may change without notice.
- Jobs can fail, stall, or return partial output. Retry logic is your responsibility.
- Beta credits can be invalidated without notice — e.g. for infrastructure migration, key compromise, or pricing reset. Beta credits do not convert to cash.
- We will give 7 days' written notice before any pricing change that affects active keys, and honor unused credits at the old rate for 7 days after the notice.
- Not suitable for production use cases where transcription failure has legal, safety, or financial consequences.
- By installing and using this package, you accept the full terms at whipscribe.com/terms.
uvx whipscribe-mcp
Alternatives:
pipx install whipscribe-mcp
pip install whipscribe-mcp
Requires Python 3.10+.
Add to claude_desktop_config.json:
{
"mcpServers": { "whipscribe": { "command": "uvx", "args": ["whipscribe-mcp"] } } }
Optional: set WHIPSCRIBE_API_KEY in env to unlock paid quota.
{
"mcpServers": { "whipscribe": { "command": "uvx", "args": ["whipscribe-mcp"], "env": { "WHIPSCRIBE_API_KEY": "your-key" } } } }
Restart Claude Desktop. Type "transcribe this podcast URL: …" and Claude will call the tool.
| Tool | Description |
|---|---|
transcribe_url(url, language?, diarize?, word_timestamps?) | Transcribe audio/video from any URL (YouTube, podcast feeds, direct file links). |
transcribe_file(path, language?, diarize?, word_timestamps?) | Transcribe a local audio/video file. |
get_job_status(job_id) | Check progress of a running job. |
get_transcript(job_id, format) | Fetch the transcript in txt, json, srt, vtt, or docx. |
list_recent_jobs(limit?) | Browse your recent jobs (local cache). |
All tools return a JSON object of the shape:
{
"ok": true, "job_id": "...", "status": "done", "transcript_preview": "first 300 chars...", "url_to_full": "https://whipscribe.com/view?id=...", "duration_sec": 967.3, "beta_notice": "..." }
On failure:
{
"ok": false, "error": { "code": "upload_failed", "message": "human-readable explanation", "retryable": true } }
| Variable | Required | Default | Purpose |
|---|---|---|---|
WHIPSCRIBE_API_KEY | No | — | API key (links jobs to your account and its credits; without it, every recording still gets a free preview) |
WHIPSCRIBE_API_BASE | No | https://whipscribe.com/api/v1 | Override API base URL (e.g. for staging) |
WHIPSCRIBE_MCP_TELEMETRY | No | 1 | Set to 0 to disable anonymous usage telemetry |
WHIPSCRIBE_MCP_POLL_TIMEOUT_SECONDS | No | 600 | Max seconds transcribe_url / transcribe_file waits before returning the job_id with a non-terminal status |
WHIPSCRIBE_MCP_POLL_INTERVAL_SECONDS | No | 3 | Seconds between job-status polls during transcription |
Opt-in anonymous telemetry (on by default; disable with WHIPSCRIBE_MCP_TELEMETRY=0):
What we collect (metadata only):
sha256(machine_id + salt)[:16]What we never collect:
The anonymization algorithm is in this repo (src/whipscribe_mcp/telemetry.py). Inspect it. If it's not acceptable, turn it off.
See whipscribe.com/pricing for current rates. Every recording gets a free preview, with or without an API key; credit packs start at $6 and never expire, and the Workspace plan is $19 a month (fair use 300 hours a month).
Apache License 2.0. Copyright 2026 Neugence Technology Pvt. Ltd.