Automated Weather Updates with Signal and FastAPI

A Signal-based assistant integrated with a structured weather service for automated weather updates and queries. This project combines FastAPI, LangChain, and Docker to deliver weather information via Signal messaging.

Features

• FastAPI-based weather-service providing daily weather forecasts and JSON responses.
• Signal assistant (assistant-core) capable of parsing direct weather commands and natural language queries.
• Structured LangChain agent for handling weather queries.
• Background cron job for daily weather notifications.
• Dockerized microservices communicating over a shared Docker network.

Tech Stack

• Python
• FastAPI
• LangChain (structured chat agent)
• Docker
• Signal API (for messaging integration)

Getting Started

Prerequisites

• Docker and Docker Compose installed
• Access to Signal API credentials/configuration

Installation & Running

1. Clone the repository:

git clone https://github.com/justin-napolitano/signal-weather-stack-tool.git
cd signal-weather-stack-tool

2. Build and start the Docker containers:

docker-compose up --build

3. The services will run on the assistant-net Docker network, with:

   • weather-service exposing /today?city=&state= endpoint
   • assistant-core exposing /inbox for Signal message handling

4. Configure environment variables such as CRON_SCHEDULE and NOTIFY_URL as needed.

Project Structure

• index.md: Project documentation and overview.
• weather-service: FastAPI app providing weather data and daily forecast notifications.
• assistant-core: Signal message handler that routes weather queries and sends responses.
• notifier-gateway: HTTP gateway forwarding messages to Signal.

Future Work / Roadmap

• Complete and extend documentation and usage examples.
• Add support for more natural language commands and additional weather data.
• Improve error handling and retry mechanisms for gateway communication.
• Enhance scheduling flexibility and configuration management.
• Integrate more advanced LangChain agents or alternative LLM tools.

For detailed usage and architecture, refer to the index.md file within the repository.