{
"markdown": "\n\n# ⚡ ShortcutsBridge\n\n**Run your Apple Shortcuts from ChatGPT — on your own Mac.**\n\n[](./LICENSE)\n[](https://www.npmjs.com/package/shortcuts-agent)\n\n
\n\nShortcutsBridge is a [Model Context Protocol](https://modelcontextprotocol.io) (MCP) connector that lets ChatGPT list, search, and **run** your **Apple Shortcuts**. It never uploads your shortcuts to a third party — 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.\n\n```\n ChatGPT ──OAuth──► ShortcutsBridge relay ──job queue──► shortcuts-agent ──► Shortcuts.app\n (connector) (Vercel, stateless) (your Mac, polls & executes) (via the shortcuts CLI)\n```\n\nYour 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.\n\nBecause 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.\n\n---\n\n## For users — 3 steps\n\n**1. Create your account** at **[shortcutsbridge.vercel.app](https://shortcutsbridge.vercel.app)** and generate a pairing code.\n\n**2. Install the Mac agent** (needs [Node 18+](https://nodejs.org)):\n\n```bash\nnpx shortcuts-agent pair # link this Mac to your account\nnpx shortcuts-agent install # keep it running on every restart\n```\n\nThe 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.\n\n**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**.\n\nThat's it. In a new chat: *\"Search my Apple Shortcuts for anything about focus\"* or *\"Run my Morning Briefing shortcut.\"*\n\n> 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.\n\n### Managing the agent\n\n```bash\nnpx shortcuts-agent status # is it paired & reachable?\nnpx shortcuts-agent logs # recent activity\nnpx shortcuts-agent uninstall # stop auto-start\n```\n\n---\n\n## What ChatGPT can do\n\n| Tool | Action |\n|------|--------|\n| `list_shortcuts` | List every shortcut on your Mac (read-only) |\n| `search_shortcuts` | Find shortcuts by keyword (read-only) |\n| `list_folders` | List your Shortcuts folders (read-only) |\n| `run_shortcut` | Run a shortcut by name, with optional text input, and return its output |\n\nThe 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.\n\n---\n\n## Self-hosting\n\nYou can run your own relay so nothing depends on the hosted instance.\n\n**Prereqs:** a [Vercel](https://vercel.com) account and a [Supabase](https://supabase.com) project (free tiers are fine).\n\n1. **Storage** — in your Supabase project's SQL editor, run [`supabase/schema.sql`](./supabase/schema.sql). It creates a tiny Redis-shaped KV + queue surface (`nb_*` functions) with RLS on.\n2. **Deploy** the `server/` directory to Vercel.\n3. **Set env vars** (see [`server/.env.example`](./server/.env.example)):\n - `SUPABASE_URL`, `SUPABASE_SERVICE_ROLE_KEY` — your project + service-role key\n - `JWT_SECRET` — `openssl rand -hex 32`\n4. **Verify:** `curl https://YOUR-APP.vercel.app/api/health` → `redisConfigured`, `redisOk`, `jwtSecretSet` all `true`.\n5. Point the agent at it: `npx shortcuts-agent pair --server https://YOUR-APP.vercel.app`.\n\n---\n\n## How it works\n\n- **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`).\n- **Relay** — MCP tool call → job pushed to `jobs:` (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.\n- **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.\n\n## Repository layout\n\n```\nserver/ Vercel functions: OAuth + MCP endpoint + relay (deploy this)\nagent/ shortcuts-agent — the npm-published Mac CLI (users npx this)\nkit/ Browser automations: register the dev-mode connector AND\n fill the OpenAI directory submission (see kit/README.md)\nsupabase/ schema.sql for self-hosting the storage\ntest/ end-to-end OAuth + MCP integration test\n```\n\n## Development\n\n```bash\n# server\ncd server && cp .env.example .env.local # fill in, then: vercel dev\n# end-to-end test against a live deployment (reads ../.env.local)\nnode test/e2e-oauth.mjs\n# agent unit tests\nnode --test agent/test-agent-unit.mjs\n```\n\n## Security & privacy\n\n- Shortcuts are listed and run **only on your Mac**. The relay never persists shortcut content — job payloads live in Redis-shaped storage with a short TTL and are deleted on delivery.\n- The agent token is stored at `~/.shortcutsbridge-agent.json` (mode `600`).\n- No secrets are committed; see [`.gitignore`](./.gitignore) and the `.env.example` files.\n- `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.\n\n## License\n\n[MIT](./LICENSE) © Isaiah Dupree\n",
"bytes": 6548,
"sha": "eac741023546d74a9cf8849574b89312e465981d618fdaf4633b17b2256a2e6c",
"repo_slug": "isaiahdupree/shortcuts-bridge",
"fonte": "repo",
"truncated": false,
"api": "https://agentalog.com/api/listings/mcp_io_github_isaiahdupree_shortcuts_bridge_5de47b84/readme"
}