🧠 Home Assistant Log Summarizer

A proof-of-concept project that summarizes recent home-assistant.log entries using OpenAI’s GPT-4o-mini, either from within Home Assistant or via an optional Streamlit debug interface.

This repository includes:

• A Streamlit-based log viewer and LLM-powered summarizer in streamlit_app/
• A Home Assistant custom integration in custom_components/log_summarizer/

Features

•	Filters and groups warnings, errors, and critical messages
•	Summarizes recent logs using GPT-4o-mini
•	Provides actionable fixes with context
•	Lets you choose a time window (e.g. last 24h)
•	Persistent notification and optional markdown output
•	Debug mode for viewing and exporting filtered logs

Installing via HACS (Home Assistant Community Store)

You can also install this integration via HACS as a custom repository:

1. In Home Assistant, go to HACS → Integrations.
2. Click the three-dot menu in the upper right → Custom repositories.
3. Enter the repository URL: https://github.com/borsic77/HA_log_analyzer
4. Select Integration as the category and click Add.
5. You’ll now see Log Summarizer available to install under HACS → Integrations.
6. After installation, restart Home Assistant and follow the same configuration steps as below, starting from step 3.

This method ensures easier future updates and visibility through HACS.

---

Installing as a Home Assistant Custom Component (ignore steps 1 and 2 if you use HACS)

1. Copy the contents of custom_components/log_summarizer/ into your Home Assistant's config/custom_components/log_summarizer/ directory.

2. In your configuration.yaml, add the following block (you can store the API key in secrets.yaml):

log_summarizer:
  api_key: !secret openai_api_key

3. In secrets.yaml, add:

openai_api_key: sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

To get your API key, create an account and generate a key at OpenAI's API key page.

Note: Accessing the OpenAI API is separate from a ChatGPT Plus subscription. Even if you have ChatGPT Plus, you'll need to create a separate OpenAI account at platform.openai.com and set up billing.

The OpenAI API uses a pay-as-you-go prepaid model, where you're billed per token (roughly per word). As of April 2025, the pricing is:

• GPT-4o: $0.005 / 1K input tokens, $0.015 / 1K output tokens
• GPT-3.5-turbo: $0.0005 / 1K input tokens, $0.0015 / 1K output tokens

This integration is designed to be run manually (e.g., on demand, not constantly), so usage costs are typically very low — often just a few cents per run.

4. Restart Home Assistant.

5. Use Developer Tools → Actions to call the log_summarizer.summarize_logs service and optionally provide the file_path and model.

---

Optional: Run in Debug Mode with Streamlit

1. Clone the repository

git clone https://github.com/your-username/HA_log_analyzer.git
cd HA_log_analyzer

2. Set up the virtual environment (using uv)

uv venv
source .venv/bin/activate
uv pip install -e .

3. Add your OpenAI API key

Create a file called .env in the project root with the following contents:

OPENAI_API_KEY=sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

4. Run the app

streamlit run streamlit_app/main.py

---

Created by Boris Legradic · MIT License