Pace Privacy Policy

Last updated: 2026-08-28

Pace is an alarm app for iPhone. It is built so that your data stays on your phone. The one exception is a feature you turn on yourself: connecting an AI agent. This page explains exactly what that sends, where it goes, and how to undo it.

What Pace stores on your phone

Pace 1.0 sends nothing

Pace 1.0 makes no network requests. Nothing leaves the phone.

Pace 1.1: connecting an AI agent (optional)

Pace 1.1 lets you connect an AI agent (for example Claude or Codex) so it can see your upcoming wakes and propose changes. This is off until you open Settings → Connect an agent and enter the pairing code the app shows into the agent. Until you do that, 1.1 behaves exactly like 1.0: no network requests.

What is sent, and where

Once an agent is paired, the app sends the following to the Pace sync server (mcp-production-e0af.up.railway.app, operated by us, hosted on Railway):

The server also stores the changes an agent proposes until your phone applies or rejects them, and which agents are connected.

Stored under a pseudonymous device identity

There is no account. When the app first contacts the server it receives a random device identifier and a random device token, kept in the iPhone keychain. The server stores only a hash of that token. We do not ask for, and the server never receives, your name, email address, phone number, Apple ID, location, contacts, or advertising identifier. Nothing stored on the server links the data to you as a person; it is linked only to the random identifier your phone holds.

Who can read it

Only AI agents you connected yourself, using the pairing code from your phone. Each agent gets its own token, bound to your device's identity, and can read only that data. We do not read it, sell it, share it, or use it for anything other than serving your own connected agents. There are no analytics, no advertising, no crash reporting, and no tracking on the server or in the app.

The agents you connect are run by their own providers. Whatever your agent asks the Pace server for (your upcoming wakes, for example) becomes part of that agent's conversation and is handled under that provider's policy:

How long it is kept

How to see what the server holds

Settings → Connected agents → What Pace knows shows the server's copy of your data, exactly as stored. The same data is available as JSON from GET /v1/devices/me/export with your device token.

How to delete it

Deleting your last device hard-deletes everything stored under it: alarms, overrides, proposed changes, pairing codes, and every connected agent's tokens. Connected agents lose access at that moment. Deleting the app without unpairing leaves the server copy until the one-year housekeeping above removes it; if you want it gone now, unpair first.

Permissions Pace asks for, and why

PermissionUsed forLeaves the phone?
Alarms (AlarmKit)Ringing the system alarm through Silent and FocusNo
CalendarReading today's events to suggest a wake time before your first meeting. Read only, on demand, never stored beyond the suggestion shownNo
CameraPhoto and QR dismiss missions. Images and codes are checked on the phone and discardedNo
Photo libraryFallback for the photo mission when the camera is unavailableNo
MotionCounting shakes, steps, and squats during a dismiss missionNo
Media libraryPlaying a song you choose from your library when Wake Mode opensNo

You can decline any permission other than Alarms; the related feature is simply unavailable. Alarms are required for the app to ring.

What Pace does not do

Optional self-hosted server

The Pace repository is open source and includes the server described above. You can run your own copy and point the app at it; then nothing reaches ours. The Google Calendar endpoints in that server are for self-hosting only and are not used by the App Store app.

Deleting your data

Unpair (Settings → Unpair) if you connected an agent, then delete the app. Everything Pace stored is removed with it.

Contact

storminggalaxys3@gmail.com