Shortcuts Bridge

io.github.IsaiahDupree由 GitHub 账号 IsaiahDupree 发布v1.0.0更新于 Oct 24, 2026

MCP connector that lets ChatGPT list, search, and run your Apple Shortcuts via a local Mac agent

已验证Streamable HTTP可网页运行效率与工作流开发者工具

概览

AI 生成的概览

远程 MCP 连接器,让 ChatGPT 通过本机 Mac 上的小型代理列出、搜索并运行你的 Apple 快捷指令。

可以做什么
  • 列出并搜索 Mac 上已安装的快捷指令,以及列出快捷指令文件夹(只读查询)
  • 按名称运行快捷指令,可传入文本输入,并返回其输出
  • 在对话中触发你已经搭建好的任意自动化流程
  • 管理本机代理:查看配对与连通状态、查看近期活动、停止开机自启
适用场景
适合希望从 ChatGPT 触发已有 Apple 快捷指令的场景,例如晨间简报、专注模式设置,或驱动应用、调用网络 API 的脚本,而快捷指令本身始终留在你的 Mac 上。
运行要求
远程 MCP 端点,需 OAuth 登录(在托管站点创建账户并生成配对码)。一台保持唤醒的 Mac,需 Node 18+ 以通过 npx 运行本机代理,代理需与账户配对并保持运行(可选安装为登录时自启的 LaunchAgent)。自托管需要 Vercel 账户和 Supabase 项目,并设置 SUPABASE_URL、SUPABASE_SERVICE_ROLE_KEY 和 JWT_SECRET。
安装前请注意
run_shortcut 属于开放世界操作:快捷指令可控制 HomeKit、驱动应用、运行脚本或调用网络 API,因此能在你的 Mac 上执行真实操作;只能运行你已创建的快捷指令,客户端应先向你确认。每个快捷指令首次运行时 macOS 会请求授权。中继会临时存储任务负载与结果,代理令牌保存在本地文件中。自托管需要 Supabase 服务角色密钥,该密钥权限极高。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Shortcuts Bridge,将其添加到工作区。
  2. 连接 所需的 凭据。它们会被私密存储,绝不会在公开页面上展示。
  3. 为需要使用其工具的对话启用该服务。

Web executable,通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体,或 SourceWeft 对话中。智能体会读取该服务器的安装指南,先向你展示发布者、端点和登录方式,经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/io.github.IsaiahDupree%2Fshortcuts-bridge/install.md,并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前,先告诉我它的发布者、端点和登录方式,等我确认。不要把 API 密钥写进命令或我会提交的文件;问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。 客户端会在首次连接时完成 OAuth 登录。

{
  "mcpServers": {
    "shortcuts-bridge": {
      "type": "http",
      "url": "https://shortcutsbridge.vercel.app/mcp"
    }
  }
}

README

⚡ ShortcutsBridge

Run your Apple Shortcuts from ChatGPT — on your own Mac.

[License: MIT] [npm]

ShortcutsBridge is a Model Context Protocol (MCP) connector that lets ChatGPT list, search, and run your Apple Shortcuts. Reads execute locally and requested results pass through the authenticated relay — every action runs on your own Mac through a tiny local agent. The cloud piece is only a stateless relay that shuttles requests between ChatGPT and your Mac.

  ChatGPT  ──OAuth──►  ShortcutsBridge relay  ──job queue──►  shortcuts-agent  ──►  Shortcuts.app (connector)          (Vercel, stateless)      (your Mac, polls & executes)     (via the shortcuts CLI)

Your Mac is the only place your shortcuts are ever listed or run. The relay stores only short-lived job payloads and your account record; it never sees a shortcut unless a job is in flight, and jobs expire in seconds.

Because a Shortcut can do almost anything on a Mac or iPhone — control HomeKit, drive apps, run scripts, call web APIs — run_shortcut turns ChatGPT into a trigger for any automation you've already built.


For users — 3 steps

1. Create your account at shortcutsbridge.vercel.app and generate a pairing code.

2. Install the Mac agent (needs Node 18+):

bash
npx shortcuts-agent pair <YOUR-CODE>   # link this Mac to your accountnpx shortcuts-agent install            # keep it running on every restart

The first time a given shortcut runs, macOS may ask you to allow it — click OK (once per shortcut). It's per-shortcut consent, not a blanket automation grant.

3. Connect ChatGPT: Settings → Plugins → turn on Developer mode (Security & login) → Create a custom plugin → paste the MCP Server URL https://shortcutsbridge.vercel.app/mcp → Authentication OAuth → sign in & Allow.

That's it. In a new chat: "Search my Apple Shortcuts for anything about focus" or "Run my Morning Briefing shortcut."

Your Mac must be awake with the agent running. npx shortcuts-agent install makes it start automatically at login and restart if it ever crashes.

Managing the agent

bash
npx shortcuts-agent status      # is it paired & reachable?npx shortcuts-agent logs        # recent activitynpx shortcuts-agent uninstall   # stop auto-start

What ChatGPT can do

ToolAction
list_shortcutsList every shortcut on your Mac (read-only)
search_shortcutsFind shortcuts by keyword (read-only)
list_foldersList your Shortcuts folders (read-only)
run_shortcutRun a shortcut by name, with optional text input, and return its output

The first three are read-only catalogue lookups. run_shortcut is the only tool that does something — it's open-world (the shortcut can take real-world actions), so ChatGPT confirms with you before running one, and only shortcuts you've already built can run.


Self-hosting

You can run your own relay so nothing depends on the hosted instance.

Prereqs: a Vercel account and a Supabase project (free tiers are fine).

  1. Storage — in your Supabase project's SQL editor, run supabase/schema.sql. It creates a tiny Redis-shaped KV + queue surface (nb_* functions) with RLS on.
  2. Deploy the server/ directory to Vercel.
  3. Set env vars (see server/.env.example):
    • SUPABASE_URL, SUPABASE_SERVICE_ROLE_KEY — your project + service-role key
    • JWT_SECRET — openssl rand -hex 32
  4. Verify: curl https://YOUR-APP.vercel.app/api/health → redisConfigured, redisOk, jwtSecretSet all true.
  5. Point the agent at it: npx shortcuts-agent pair <CODE> --server https://YOUR-APP.vercel.app.

How it works

  • Auth — OAuth 2.1 with PKCE and Dynamic Client Registration (RFC 7591). ChatGPT registers itself, the user signs in on the ShortcutsBridge consent page, and ChatGPT gets a scoped MCP access token. JWTs are HMAC-signed; three kinds (session, agent, mcp) with distinct privileges. Optional email verification via Resend (RESEND_API_KEY); off unless configured, and enforcement is opt-in (REQUIRE_EMAIL_VERIFICATION).
  • Relay — MCP tool call → job pushed to jobs:<user> (TTL ≤ the caller's wait window, so an abandoned job can't run late) → the Mac agent (holding a hanging long-poll) is handed the job within ~200ms, executes it, pushes the result → the MCP handler returns it. The long-poll means jobs are delivered on arrival rather than on a fixed interval, so relay overhead is ~300ms instead of ~1s. The agent is considered "online" only while it's actively connected.
  • Agent — a dependency-free Node CLI. Tools shell out to the macOS shortcuts command-line tool (shortcuts list, shortcuts run); arguments are passed as argv (never shell-interpolated). A macOS LaunchAgent keeps it alive across logins and crashes.

Repository layout

server/    Vercel functions: OAuth + MCP endpoint + relay        (deploy this)agent/     shortcuts-agent — the npm-published Mac CLI          (users npx this)kit/       Browser automations: register the dev-mode connector AND           fill the OpenAI directory submission (see kit/README.md)supabase/  schema.sql for self-hosting the storagetest/      end-to-end OAuth + MCP integration test

Development

bash
# servercd server && cp .env.example .env.local   # fill in, then: vercel dev# end-to-end test against a live deployment (reads ../.env.local)node test/e2e-oauth.mjs# agent unit testsnode --test agent/test-agent-unit.mjs

Security & privacy

  • Shortcuts are listed and run only on your Mac. The relay temporarily stores requested content in short-lived job/result queues — job payloads live in Redis-shaped storage with a short TTL and are deleted on delivery.
  • The agent token is stored at ~/.shortcutsbridge-agent.json (mode 600).
  • No secrets are committed; see .gitignore and the .env.example files.
  • run_shortcut is marked open-world — a shortcut can take real-world actions — so ChatGPT asks before running one. Only shortcuts you've already built can run.

License

MIT © Isaiah Dupree

Production storage and owner verification

Relay 1.4 uses isolated service-role-only tables/functions on the configured shared Supabase project. Production refuses missing storage configuration, bounds RPC requests to ten seconds, and returns HTTP 503 when storage or signing configuration is unavailable. Apply the checked-in isolated-storage migration before deploying this version; it does not replace another bridge's account tables or functions.

node --env-file=.env.local test/live-storage.mjs verifies the actual persistent store, FIFO, expiry and counters using temporary integration keys. BRIDGE_SESSION_FILE=/private/owner-session.json BRIDGE_READ_TOOL=READ_TOOL node test/live-relay.mjs verifies the actual owner session, PKCE OAuth, hosted MCP and paired-Mac read; choose this bridge's inventory tool. No reviewer data counts as a live acceptance. Credentials and sessions belong outside Git.

来源:README.md,提交 709b687

工具

0
该服务器需要登录,因此 SourceWeft 无法在没有账号的情况下读取其工具,也未收录。已连接账号所能看到的内容可能因账号及其权限而异。

版本历史

1
  1. v1.0.0最新Sep 16, 2026

评分

已满足适用检查项的 85%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分
  • 已声明认证方式必需10 分
  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分