vibelogin-mcp

vibelogin-mcp

Add authentication to your app without leaving your IDE. Creates projects, configures auth methods, wires up OAuth, and scaffolds sign-in flows - all from your editor's chat.

Category
访问服务器

README

@vibelogin/mcp

Add authentication to your app without leaving your IDE. MCP server for Cursor, Claude Code, Windsurf, and Cline.

Just say "add authentication to my app" — the agent creates your project, configures auth methods, wires up Google OAuth, and scaffolds a working sign-in flow into your codebase. All from your editor's chat.


Install

No global install needed. Wire it into your MCP client config once and it runs on demand via bunx.


Client configuration

One snippet works in every client. Paste this into your MCP settings and you're done.

{
  "mcpServers": {
    "vibelogin": {
      "command": "bunx",
      "args": ["@vibelogin/mcp"]
    }
  }
}

The first tool call opens a browser for one-click consent. After that, tokens are cached under ~/.vibelogin/credentials.json — you won't be asked to log in again for 30 days.

Claude Desktop / Claude Code

~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "vibelogin": {
      "command": "bunx",
      "args": ["@vibelogin/mcp"]
    }
  }
}

Cursor

~/.cursor/mcp.json (or per-project .cursor/mcp.json):

{
  "mcpServers": {
    "vibelogin": {
      "command": "bunx",
      "args": ["@vibelogin/mcp"]
    }
  }
}

Windsurf / Cline / Zed

Same shape — every MCP client accepts a command + args block. The one-liner above works in all of them.

That's it — no environment variables needed. Defaults point to production automatically.


Authentication

On first use, a browser window opens for one-click consent. After that, you're authenticated for 30 days — no further prompts.


Tools

The agent picks the right tools automatically based on your conversation. Just say what you need.

create_project

Creates a new VibeLogin project in your org. Returns the project's id, slug, oauthSlug, environment, publishable key, and secret key (shown once).

Param Type Required Description
name string yes Human-readable project name.
slug string no URL-safe slug; auto-generated from name if omitted.
environment enum no production | development | staging (default production).

list_projects

Lists every project in your org. Use this when the agent needs to discover what already exists before creating something new.

No parameters.

get_project

Look up a single project by id, slug, or oauthSlug.

Param Type Required Description
idOrSlug string yes Project id, slug, or oauthSlug.

configure_auth

Toggle auth methods, set redirect URLs, and/or wire up Google OAuth. Field-level updates — anything you don't set is left untouched.

Param Type Required Description
projectId string yes The project's id.
methods.emailPassword boolean no Enable email + password sign-in.
methods.magicLink boolean no Enable magic link emails.
methods.emailOtp boolean no Enable 6-digit email OTP codes.
methods.passwordReset boolean no Enable password reset emails.
methods.emailVerification boolean no Require email verification on signup.
google.clientId string with google Google OAuth client ID (from Google Cloud Console).
google.clientSecret string with google Google OAuth client secret.
google.enabled boolean no Default true.
redirectUrls string[] no Overwrites the project's allowed redirect URL list.

If you call this without google, the response includes the exact Google Cloud Console URL + redirect URI you need to set up the credentials.

add_auth_to_project

Scaffold a working sign-in flow into your codebase. Detects the framework from package.json and writes non-destructive files (existing files are skipped, never clobbered).

Supported frameworks today: Next.js (App Router), Vite + React. Refused with guidance: Remix, Astro, Express, unknown.

Param Type Required Description
projectDir string yes Absolute path to the project root.
slug string yes The project's oauthSlug (from create_project).
publishableKey string yes The project's publishable key.
appUrl string yes Your app's origin (e.g. https://myapp.com).

Files written for Next.js

middleware.ts                       # hostedAuthMiddleware
app/auth/callback/route.ts          # createCallbackHandler
app/login/page.tsx                  # <VibeLogin />
.env.local.example                  # publishable key + secret placeholder

Next steps printed by the tool:

  1. pnpm add @vibelogin/nextjs
  2. Copy .env.local.example → .env.local, fill VIBELOGIN_SECRET_KEY
  3. pnpm dev → open /login

Files written for Vite + React

src/auth/VibeLoginProvider.tsx      # context + useVibeLogin() hook
src/auth/Login.tsx                  # /login redirect to hosted UI
src/auth/Callback.tsx               # /auth/callback handler
.env.example                        # VITE_-prefixed publishable key

Next steps printed by the tool:

  1. pnpm add @vibelogin/react react-router-dom
  2. Wrap <App/> in <VibeLoginProvider/>
  3. Add routes /login and /auth/callback
  4. Copy .env.example → .env.local
  5. pnpm dev → open /login

Example conversations

You: "Add VibeLogin to this Next.js app. Email + password and Google sign-in."

Agent: (calls create_project { name: "myapp", environment: "production" }) Created myapp (pk_live_…). (calls configure_auth { projectId, methods: { emailPassword: true } }) Email + password enabled. To turn on Google, I need a Client ID and Secret from Google Cloud Console — here's the redirect URI to authorize: https://api.vibelogin.com/oauth/myapp-xyz/google/callback. (after you paste creds:) (calls configure_auth { google: { clientId, clientSecret } }) (calls add_auth_to_project { projectDir: ".", slug: "myapp-xyz", publishableKey: "pk_live_…", appUrl: "https://myapp.com" }) Scaffold complete (next). Wrote middleware.ts, app/auth/callback/route.ts, app/login/page.tsx, .env.local.example. Run pnpm add @vibelogin/nextjs && pnpm dev and open /login.

You: "What projects do I have already?"

Agent: (calls list_projects) You have 3: myapp (production), myapp-staging (staging), internal-tools (development).


Security

  • No secrets in the agent — authentication uses loopback OAuth + PKCE (RFC 8252)
  • Single-use authorization codes with 2-minute TTL
  • Rotated refresh tokens — every refresh invalidates the previous one
  • Credentials stored securely — chmod 600 on ~/.vibelogin/credentials.json

Supported frameworks

Framework Status
Next.js (App Router) Fully supported
Vite + React Fully supported
Remix, Astro, Express Coming soon

License

Apache-2.0

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