Skip to Content
User guidePrivacy

Privacy

seedypod makes no outbound connections unless you configure one. There is no account, no telemetry and no crash reporting.

What stays where

  • Sensor history, schedules, grow cycles and pod notes live in a SQLite database on the machine that runs core.
  • Camera stills and timelapses stay on the unit, or on the core host you chose, and are served only on your LAN.
  • MQTT traffic goes only to brokers you configure, such as your Home Assistant broker.
  • The live demo  runs a simulator in your browser and talks to no backend.
  • Off-device storage is opt-in. With an S3-compatible bucket saved in Settings → Off-device storage and analysis, new stills are copied to it, and a backup of the unit goes to backups/ there every night. The bucket is yours, and seedypod talks to no other service.
  • A backup download goes only to the browser that asked for it. A backup holds the unit’s secrets (bucket keys, MQTT password), so keep it somewhere private.

Who can reach it

Core trusts your local network. By default the web UI, REST API, MCP endpoint and live event stream have no authentication, so do not port-forward or publicly reverse-proxy port 3000.

If you want more than LAN trust, set an API key in Settings → API access. Clients that are not on the unit itself must then send it as Authorization: Bearer YOUR_KEY or in an x-api-key header. Requests from the unit itself (its own UI, scripts and schedules) are always allowed.

MQTT credentials you enter for the Home Assistant bridge are stored in core’s database in plain text: anyone who can read that database already controls the unit.

This website

This documentation site is a static site on GitHub Pages. It sets no cookies of its own and has no analytics. Search runs in your browser.

Source: docs/PRIVACY.md · ADR 0012: Deployment and trust model

Last updated on
DebuggingiPhone app