Skip to content

Installation

TendrilGrow is distributed as a HACS custom integration.

Prerequisites

  • Home Assistant 2026.2.0 or newer.
  • HACS installed and configured.
  • Any companion integrations that expose the entities you want to map, for example:
    • A controller integration for lights/fans (the maintainer uses the Vivosun HACS integration).
    • The Tuya HACS integration for reservoir water monitors (optional).
    • A camera integration (required for AI vision health checks).

No hardcoded entities

TendrilGrow never assumes specific entity IDs. You map your own entities during setup, so any brand of sensor, controller, or camera works.

Install via HACS

  1. Open HACS in Home Assistant.
  2. Open the menu (top-right) and choose Custom repositories.
  3. Add the repository URL https://github.com/Trec-TorConsulting/TendrilGrow.
  4. Select category Integration.
  5. Install TendrilGrow from HACS.
  6. Restart Home Assistant. Custom integrations are not loaded until a restart.
  7. Go to Settings → Devices & Services → Add Integration and add TendrilGrow.

Manual installation

If you prefer not to use HACS, copy the integration folder into your Home Assistant configuration:

/config/custom_components/tendrilgrow

Then restart Home Assistant and add the integration from Settings → Devices & Services. A packaged tendrilgrow.zip is attached to each GitHub release.

First-time setup

Add one integration entry per grow space (tent, room, or zone). The setup flow walks you through:

  1. Grow-space name and type.
  2. Mapping sensor and control entities (all mappings are optional). You can also enable Tuya cloud polling here.
  3. AI health options (check interval, critical-score threshold, optional notify service, result retention).
  4. AI provider selection (None, Gemini, OpenAI, or Ollama).
  5. Provider credentials/endpoint.
  6. Model selection from discovered models (with a manual fallback).

See Configuration for a detailed walkthrough of every step and field.

To enable AI vision checks

Map a camera entity and select a vision-capable provider and model. Checks then run on a schedule, on demand from the run button, or via the tendrilgrow.run_ai_health_check service.