<!-- Generated from skills/deployment.md by tools/generate_integrations.py. Edit the source file, not this one. -->

Skill: Deployment

Purpose

Deploy applications, expose services, and manage release workflows.

Tools Required

• Deployment CLI tools (Netlify, Vercel, gh-pages, etc.)
• Shell commands
• Port exposure tools

General Principles

• Read deployment config before attempting to deploy
• Ensure all required files exist before deployment
• Verify build succeeds before deploy
• Check deployment status after deploying
• NEVER auto-deploy to production without user confirmation

Deployment Workflow

Pre-Deployment Checklist

• Read deployment configuration (if exists)
• Verify all source files are present
• Build succeeds locally
• Environment variables configured
• Required service accounts / API keys available

Common Platforms

Platform	Tool	Setup
GitHub Pages	gh-pages package or Actions	Push to gh-pages branch
Netlify	Netlify CLI / manual	Connect repo + configure build
Vercel	Vercel CLI	vercel for project setup
Static hosting	rsync / scp / s3	Copy build output to host

Deployment Steps

1. Build the project
2. If deploying new site: use project_id empty
3. If updating existing site: use existing project_id
4. Run deploy command
5. Check deployment status
6. Provide access URL to user

Service Exposure

• For temporary services: expose local ports
• For permanent services: deploy to cloud hosting
• Provide access links after deployment
• Monitor deployed applications

PR/MR Workflow

For implementation tasks in collaborative projects:

1. Create feature branch from updated main
2. Make changes
3. Commit with descriptive message
4. Push branch
5. Create Pull/Merge Request
6. Include summary of changes in PR description

Environment Configuration

• Use environment variables for config
• Provide .env.example file with placeholder values
• Document required environment variables
• Never commit real secrets to version control

Rollback

• Know how to roll back a deployment
• For git-based deploys: revert commit and redeploy
• For platform deploys: use platform's rollback feature
• Always keep previous deployment artifacts