Help My Pet

An AI-powered Telegram bot that helps pet owners by answering questions about pet health using the Anthropic Claude AI model.

Features

• Answers pet health-related questions using Anthropic's Claude AI
• Provides helpful and compassionate advice
• Always recommends consulting with a veterinarian for serious medical conditions
• Easy to use through Telegram messaging

Prerequisites

• Go 1.21 or later
• Telegram Bot Token (obtain from @BotFather)
• Anthropic API Key (obtain from Anthropic Console)

Installation

1. Clone the repository:

git clone https://github.com/ksysoev/help-my-pet.git
cd help-my-pet

2. Install dependencies:

go mod download

3. Create a configuration file:

cp config.yaml config.local.yaml

4. Edit config.local.yaml and add your API keys:

bot:
  telegram_token: "your-telegram-bot-token" # Required: Your Telegram bot token from BotFather
  rate_limit:
    user_hourly_limit: 5   # Maximum number of requests per hour per user
    user_daily_limit: 15   # Maximum number of requests per day per user
    global_daily_limit: 1000  # Maximum total requests per day across all users
    whitelist_ids: []      # List of user IDs exempt from rate limiting

ai:
  model: "claude-2"      # Optional: Anthropic model to use (default: claude-2)
  api_key: ""           # Required: Your Anthropic API key
  max_tokens: 1000      # Optional: Maximum tokens in response (default: 1000)

Usage

1. Start the bot:

go run cmd/help-my-pet/main.go bot --config config.local.yaml

2. Open Telegram and start chatting with your bot
3. Ask any pet health-related questions

Development

• Run tests:

go test ./...

• Run linter:

golangci-lint run

Docker

You can run the bot using Docker in the following ways:

Pull from GitHub Packages

1. Pull the image:

docker pull ghcr.io/ksysoev/help-my-pet:latest

2. Create your config file as described in the Installation section

3. Run the container:

docker run -v $(pwd)/config.local.yaml:/app/config.yaml ghcr.io/ksysoev/help-my-pet bot

Build Locally

1. Build the image:

docker build -t help-my-pet .

2. Run the container:

docker run -v $(pwd)/config.local.yaml:/app/config.yaml help-my-pet bot

Deployment

The application is deployed to production using GitHub Actions and the cloudlab deployment workflow.

Prerequisites

• Docker Swarm initialized on the cloudlab server
• GitHub Actions secrets configured
• Bot images published to GitHub Container Registry (ghcr.io)

Required GitHub Secrets

Set up the following secrets in your GitHub repository (Settings → Secrets and variables → Actions):

SSH Connection:

• DO_SSH_PRIVATE_KEY - SSH private key for accessing the cloudlab server
• DO_HOST - Cloudlab server IP address (e.g., 167.172.190.133)
• DO_USER - SSH username for deployment (e.g., deployer)

Application Secrets:

• TELEGRAM_TOKEN - Telegram bot token from @BotFather
• ANTHROPIC_API_KEY - Anthropic API key from Anthropic Console
• CODECOV_TOKEN - Codecov token for coverage reports

Environment Variable Mapping:

• TELEGRAM_TOKEN (GitHub secret) → BOT_TELEGRAM_TOKEN (container)
• ANTHROPIC_API_KEY (GitHub secret) → AI_API_KEY (container)

Deployment Process

The deployment is fully automated via GitHub Actions:

1. On Pull Request: Build and test the Docker image (amd64 only)
2. On Tag Push (e.g., v1.0.0):
   • Build multi-arch Docker images (amd64, arm64)
   • Push images to GitHub Container Registry
   • Deploy to production using the cloudlab workflow
   • Automatic rollback on failure

To deploy a new version:

git tag v1.0.0
git push origin v1.0.0

Manual deployment (if needed):

# SSH to the cloudlab server
ssh -p 1923 deployer@167.172.190.133

# Navigate to deployment directory
cd ~/cloudlab/stacks/helpmypet

# Deploy the stack
docker stack deploy -c docker-compose.yml helpmypet

Monitoring

Check service status:

ssh -p 1923 deployer@167.172.190.133
docker stack ps helpmypet

View service logs:

# Bot logs
docker service logs -f helpmypet_help-my-pet

# Redis logs
docker service logs -f helpmypet_redis

Health check:

# Check if services are running
docker service ls | grep helpmypet

Rollback

If deployment fails, the system automatically rolls back to the previous version. For manual rollback:

docker service update --rollback helpmypet_help-my-pet

Project Structure

• cmd/help-my-pet/ - Main application entry point
• pkg/
  • bot/ - Telegram bot service implementation
  • core/ - Core business logic and AI service
  • cmd/ - Command line interface and configuration

License

This project is licensed under the MIT License - see the LICENSE file for details.