/linebot-openai-agent

LINEBot template with OpenAI Agent SDK but using Google Gemini custom provider.

Primary LanguagePython

LINE Bot with OpenAI Agent and Google Gemini

Project Background

This project is a LINE bot that uses both OpenAI Agent functionality and Google Gemini models to generate responses to text inputs. The bot can answer questions in Traditional Chinese and provide helpful information.

Screenshot

image

Features

  • Text message processing using OpenAI Agent in Traditional Chinese
  • Support for function calling with tools like weather information and translation
  • Integration with LINE Messaging API for easy mobile access
  • Built with FastAPI for efficient asynchronous processing

Technologies Used

  • Python 3
  • FastAPI
  • LINE Messaging API
  • OpenAI Agent framework
  • Aiohttp
  • PIL (Python Imaging Library)

Setup

  1. Clone the repository to your local machine.

  2. Set the following environment variables:

    • ChannelSecret: Your LINE channel secret
    • ChannelAccessToken: Your LINE channel access token
    • EXAMPLE_BASE_URL: Your OpenAI compatible API base URL. If you want to use Google Gemini, you should fill https://generativelanguage.googleapis.com/v1beta/ here.
    • EXAMPLE_API_KEY: Your OpenAI compatible API key, If you want to use Gemini
    • EXAMPLE_MODEL_NAME: The model name to use with the API (e.g., "gpt-4")

  3. Install the required dependencies:

    pip install -r requirements.txt

  4. Start the FastAPI server:

    uvicorn main:app --reload

  5. Set up your LINE bot webhook URL to point to your server's endpoint.

Usage

Text Processing

Send any text message to the LINE bot, and it will use the OpenAI Agent to generate a response in Traditional Chinese.

Available Tools

The bot has access to the following function tools:

  • get_weather: Get weather information for a specified city
  • translate_to_chinese: Translate text to Traditional Chinese

Deployment Options

Local Development

Use ngrok or similar tools to expose your local server to the internet for webhook access:

ngrok http 8000

Docker Deployment

You can use the included Dockerfile to build and deploy the application:

docker build -t linebot-openai-agent .
docker run -p 8000:8000 -e ChannelSecret=YOUR_SECRET -e ChannelAccessToken=YOUR_TOKEN -e EXAMPLE_BASE_URL=YOUR_BASE_URL -e EXAMPLE_API_KEY=YOUR_API_KEY -e EXAMPLE_MODEL_NAME=YOUR_MODEL linebot-openai-agent

Cloud Deployment

  1. Install the Google Cloud SDK and authenticate with your Google Cloud account.

  2. Build the Docker image:

    gcloud builds submit --tag gcr.io/$GOOGLE_PROJECT_ID/linebot-openai-agent

  3. Deploy the Docker image to Cloud Run:

    gcloud run deploy linebot-openai-agent --image gcr.io/$GOOGLE_PROJECT_ID/linebot-openai-agent --platform managed --region $REGION --allow-unauthenticated

  4. Set up your LINE bot webhook URL to point to the Cloud Run service URL.