mcp-wormhole

mcp-wormhole

Open-source MCP servers that connect AI agents to tools like Asana and Vercel via natural language.

Category
访问服务器

README

<div align="center">

mcp-wormhole

Open-source MCP servers that connect AI agents to the tools you already use.

Website npm @mcp-wormhole/asana npm @mcp-wormhole/vercel License: MIT MCP

Documentation · Blog · Contributing · GitHub

mcp-wormhole server catalog — orange CLI demo

</div>


What is this?

mcp-wormhole is a monorepo of Model Context Protocol (MCP) servers — one package per integration. Each server wraps a third-party vendor API so AI clients (Cursor, Claude Desktop, VS Code Copilot, Windsurf, and 16 others) can read and act on your tools through natural language.

No new backends. No proprietary proxies. Just stdio MCP servers published to npm.

You  →  AI Client (Cursor, Claude, …)  →  MCP Server (npx)  →  Vendor API (Asana, Slack, …)

Owner: @Ayush7614


Table of contents


Quick start

Use a published server (no clone required)

Add this to your MCP client config (Cursor: ~/.cursor/mcp.json, Claude Desktop: claude_desktop_config.json, VS Code: .vscode/mcp.json):

{
  "mcpServers": {
    "asana": {
      "command": "npx",
      "args": ["-y", "@mcp-wormhole/asana"],
      "env": {
        "ASANA_ACCESS_TOKEN": "your_token_here"
      }
    },
    "vercel": {
      "command": "npx",
      "args": ["-y", "@mcp-wormhole/vercel"],
      "env": {
        "VERCEL_TOKEN": "your_token_here"
      }
    }
  }
}

For team-scoped Vercel projects, add "VERCEL_TEAM_ID": "team_…" to the vercel env block.

Restart your client, then ask: "List my open Asana tasks" or "List my Vercel projects".

Published on npm: @mcp-wormhole/asana (0.2.0 · 66 tools · 18 prompts) · @mcp-wormhole/vercel (0.2.0 · 18 tools · 8 prompts)

Get tokens: Asana developer console · Vercel account tokens

Server guides: Asana MCP · Vercel MCP · All integrations

Clone for development

git clone https://github.com/Ayush7614/mcp-wormhole.git
cd mcp-wormhole
pnpm install
pnpm build

Available servers

Browse all: Asana · Vercel · Slack · Sentry · Google Calendar · Airtable · Stripe · Cloudflare · GitHub Actions · PagerDuty · Linear

Server npm package Status Auth Tools
Asana @mcp-wormhole/asana Available PAT 66 tools · 18 prompts · resources
Vercel @mcp-wormhole/vercel Available API token 18 tools · 8 prompts · resources
Slack @mcp-wormhole/slack Planned Bot token
Sentry @mcp-wormhole/sentry Planned Auth token
Google Calendar @mcp-wormhole/google-calendar Planned OAuth
Airtable @mcp-wormhole/airtable Planned PAT
Stripe @mcp-wormhole/stripe Planned Secret key
Cloudflare @mcp-wormhole/cloudflare Planned API token
GitHub Actions @mcp-wormhole/github-actions Planned PAT
PagerDuty @mcp-wormhole/pagerduty Planned API key
Linear @mcp-wormhole/linear Planned API key

Each server calls the vendor's existing REST API — we don't host new backends.


Connect your client

Step-by-step guides with copy-paste configs for 20 AI clients:

Client Asana Vercel
Cursor Cursor + Asana Cursor + Vercel
VS Code VS Code + Asana VS Code + Vercel
Claude Desktop Claude + Asana Claude + Vercel
Claude Code Claude Code + Asana Claude Code + Vercel
…and 16 more All integrations same page — pick Vercel server

Full server walkthroughs:


Repository structure

mcp-wormhole/
├── packages/
│   ├── asana/              # @mcp-wormhole/asana — live on npm
│   ├── vercel/             # @mcp-wormhole/vercel
│   ├── _template/          # Copy this to start a new server
│   ├── slack/              # planned
│   └── …
├── site/                   # Docs site (Vite + React, GitHub Pages)
│   ├── src/data/           # Servers, integrations, guides, blog
│   └── public/demo/        # Verification GIFs
├── package.json            # pnpm workspace root
├── pnpm-workspace.yaml
├── CONTRIBUTING.md
└── README.md

Each package is a standalone MCP server with its own package.json, tools, verify script, and README.


Development

# Install all workspace dependencies
pnpm install

# Build everything
pnpm build

# Build one package
pnpm --filter @mcp-wormhole/asana build
pnpm --filter @mcp-wormhole/vercel build

# Verify Asana server against real API
cd packages/asana
cp .env.example .env   # add ASANA_ACCESS_TOKEN
pnpm verify

# Verify Vercel server against real API
cd ../vercel
cp .env.example .env   # add VERCEL_TOKEN
pnpm verify

# Run docs site locally
cd site && npm run dev
# → http://localhost:5173

Adding a new server

  1. Copy packages/_templatepackages/<name>
  2. Implement tools against the vendor's official API (Zod validation, MCP SDK)
  3. Document env vars in README + .env.example
  4. Add entry in site/src/data/servers.ts for the docs catalog
  5. Open a PR — one server per PR

See CONTRIBUTING.md for the full checklist, tool naming conventions, and publishing workflow.


Publishing to npm

Packages publish under the @mcp-wormhole npm org.

Package Latest npm
Asana 0.2.0 @mcp-wormhole/asana
Vercel 0.2.0 @mcp-wormhole/vercel
pnpm --filter @mcp-wormhole/asana build
cd packages/asana
npm publish --access public --otp=YOUR_CODE

Or use GitHub Actions → Publish npm packages → Run workflow (requires NPM_TOKEN secret).

Details: CONTRIBUTING.md § Publishing


Guidelines

For contributors

Rule Detail
One server per PR Keep reviews focused
Vendor API only No scraping, no unofficial endpoints
Read tools first list_, get_, search_ before create_, update_, delete_
Zod everywhere Validate all tool inputs
No secrets Never commit tokens; use .env.example
Verify script Hit the real API — no mocks in pnpm verify
Update docs Add server to site/src/data/servers.ts + root README table

For users

Rule Detail
Keep tokens local MCP config env vars stay on your machine
Restart after config MCP clients load servers at startup
Use npx No repo clone needed for published packages
Report issues GitHub Issues

Stack

  • Runtime: Node.js 18+
  • Language: TypeScript
  • MCP SDK: @modelcontextprotocol/sdk
  • Validation: Zod
  • Build: tsup
  • Transport: stdio (local process via npx)

Blog

Tutorials and release notes on the docs site:


License

MIT — see LICENSE.


<div align="center">

Built by @Ayush7614

</div>

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选