OpenViSTAI · Agents · Blockchain · Crypto
Menu
Workflow Agents

Home Generative Agent

Home Generative Agent is a Python integration for Home Assistant that adds a LangGraph-based generative agent for conversational smart-home control, automation creation, camera analysis, semantic memory and Sentinel anomaly checks. It supports cloud models and local or edge models through Ollama or OpenAI-compatible servers.

Independent OpenViST project profile, based on linked sources.

Home Generative Agent — OpenViST project profile in Workflow Agents.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when Home Assistant is the center of a smart-home setup and you want an agent that can act on entities or create automations, rather than only return conversational responses.

Who it is for

  • Home Assistant users seeking natural-language control of smart-home entities.
  • Home Assistant users who want to create and register automations through chat.
  • Operators who want camera analysis or configured Sentinel alerts within their Home Assistant setup.

Documented uses

  • Ask questions about Home Assistant entities and control them conversationally.
  • Describe an automation in plain language and have the agent write and register its YAML.
  • Ask what a camera sees or enable motion-triggered analysis and anomaly detection.
  • Use semantic search across past conversations for retained preferences and context.
  • Run configured Sentinel checks and receive alerts about detected conditions.
Before you choose

Requirements & limits

The software is public on GitHub under the MIT license. HACS is the recommended installation path, with manual installation also supported. The documented minimum is Home Assistant 2026.8.0; the bundled PostgreSQL/pgvector app requires Home Assistant OS or Supervised.

  • PostgreSQL with pgvector is a required component for the documented setup.
  • A supported model provider must be configured before running Setup.
  • Local model use requires a compatible server such as Ollama or another OpenAI-compatible server; cloud providers are also listed.
  • Face recognition is optional, off by default and dependent on an external face-service.
  • Voice features are optional and require an OpenAI API or a compatible local server.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install and configure the basic integration

Add Home Generative Agent to Home Assistant and configure a model provider and setup mode.

Before you start

  • Home Assistant 2026.8.0 or later.
  • HACS for the recommended install route; manual installation is also supported.
  • PostgreSQL with pgvector, either the bundled app or a separate server.
  • At least one supported model provider.

Steps

  1. Install and configure the PostgreSQL with pgvector app. The bundled app route requires Home Assistant OS or Supervised.

  2. Install Home Generative Agent from HACS.

  3. Restart Home Assistant, then use Settings → Devices & Services → Add Integration and select Home Generative Agent.

  4. On the integration page, add a model provider and choose Edge or Cloud, then select a provider type.

  5. Click + Setup and choose a setup mode.

Expected outcome

The documented quick-start process configures the integration and its model provider in Home Assistant.

Guide limitations

  • This guide describes setup; it does not verify a working deployment.
  • The bundled database-app route is unavailable on Home Assistant Container or Core; the README points to a separate PostgreSQL option.
  • The source text names Basic setup and recommended defaults but does not include the complete remaining setup instructions.
Evidence and transparency

Sources & review

Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
goruck/home-generative-agent · Official homepage title · retained in its source language