deploy-light-bot-to-production

name	Deploy Light Bot to Production
description	Promote staging commit to production and deploy using SSH and Docker Compose.

Deploy Light Bot to Production

Promote the current staging commit to production and deploy the Light Bot application.

Server Configuration

Staging Server

• SSH Key: ~/.ssh/personal
• SSH Host: [e-mail]
• Project Directory: services/light-bot-staging
• Service Name: light-bot-staging
• Docker Compose File: /root/services/docker-compose.yml

Production Server

• SSH Key: ~/.ssh/personal
• SSH Host: [e-mail]
• Project Directory: services/light-bot
• Service Name: light-bot
• Telegram Channels: @power_po2 (same as staging but without _test suffix)
• Docker Compose File: /root/services/docker-compose.yml

Task

Execute the deployment process to promote staging to production:

Step 1: Get Staging Commit Hash

Fetch the current git commit hash from the staging server:

ssh -i ~/.ssh/personal [e-mail] "cd services/light-bot-staging && git rev-parse HEAD"

Step 2: Fast-Forward Local Prod Branch

1. Ensure local repo is clean (warn if there are uncommitted changes)
2. Fetch latest changes
3. Fast-forward local prod branch to the staging commit:

   git fetch origin
   git checkout prod
   git merge --ff-only <STAGING_COMMIT_HASH>
   

   If fast-forward fails, stop and ask the user for guidance.

Step 3: Push Prod Branch

Push the updated prod branch to remote (regular push, no force):

git push origin prod

Step 4: Check and Sync Environment Variables

Before deployment, check if there are new environment variables in .env.example (from prod branch) that need to be added to production .env:

1. Fetch both files:
   • .env.example from local prod branch (just updated)
   • .env from the production server
2. Compare them to find any new variables in .env.example that are missing in .env
3. If new variables are found:
   • Add them to .env with the default values from .env.example
   • Important: Do NOT modify existing values in .env - they may be intentionally different
   • Note: Ensure TELEGRAM_CHANNEL_ID and TELEGRAM_SCHEDULE_CHANNEL_ID are set to @power_po2 in production
   • Ask the user if unsure about any changes
4. If changes were made, upload the updated .env back to the production server

Step 5: Run Deployment Script

Once prod branch is pushed and .env is synced, run the deployment script:

./.claude/skills/deploy-prod/deploy.sh ~/.ssh/personal [e-mail] services/light-bot light-bot

The script will:

1. SSH into the production server
2. Navigate to project directory
3. Checkout and pull prod branch
4. Rebuild the Docker image
5. Restart the service with docker-compose

Step 6: Return to Main Branch

After deployment, switch back to main branch:

git checkout main