| name | devops-helper |
| description | This skill should be used when the user asks to "set up deployment", "configure CI/CD", "create a Dockerfile", "choose a hosting platform", or discusses deployment and infrastructure decisions. Provides pragmatic DevOps guidance for indie developers. |
DevOps Helper Skill
Auto-activated when: User discusses deployment, CI/CD, Docker, testing automation, or infrastructure decisions.
Core Capabilities
1. CI/CD Pipelines
- GitHub Actions workflow design
- Test automation (unit, integration, e2e)
- Build optimization
- Deployment strategies (blue/green, canary, rolling)
- Secrets management
2. Containerization
- Dockerfile best practices (multi-stage builds)
- Docker Compose for local dev
- Image optimization (layer caching, size reduction)
- Container orchestration (when needed)
3. Deployment Platforms
- Platform selection (Vercel, Render, Fly.io, Railway)
- Cost optimization strategies
- Database hosting (Supabase, PlanetScale, Neon)
- Static site hosting (Netlify, GitHub Pages)
4. Testing Automation
- Test pyramid (unit > integration > e2e)
- Pre-commit hooks (Husky, lint-staged)
- Code quality gates (coverage thresholds)
- Performance testing basics
Design Philosophy: Solid Indie
Ship Fast Principles:
- Start with platform-as-a-service (Vercel, Render)
- Use managed databases (don't self-host Postgres initially)
- Automate tests, but don't obsess over 100% coverage
- Monitor errors (Sentry), not metrics initially
Right-Sized DevOps:
- Solo/small team: GitHub Actions + Vercel
- Need scaling: Add Fly.io or Railway
- Need full control: Then consider AWS/GCP (not before)
Anti-Patterns to Avoid:
- ❌ Kubernetes for < 10 person teams
- ❌ Self-hosted infrastructure before product-market fit
- ❌ Complex multi-environment setups (dev/staging/prod initially)
- ❌ Over-monitoring (start with errors, add metrics later)
Delegation Strategy
When analysis is needed, I will:
- Quick assessment - Identify DevOps problem
- Delegate to agents when feasible:
devops-engineer agent for pipeline optimization
experienced-engineer agent for testing strategy
performance-engineer agent for build performance
- Run in background - Use Task tool with
run_in_background: true
- Synthesize results - Provide actionable DevOps recommendations
Example Activation
User: "I need to set up deployment for my Next.js app with a PostgreSQL database"
Skill activates and provides:
1. Platform recommendation (Vercel for Next.js + Supabase for PostgreSQL)
2. CI/CD setup (GitHub Actions for tests, Vercel auto-deploys)
3. Environment variables strategy (.env.local, Vercel env vars)
4. Database migrations approach (Drizzle or Prisma)
5. Delegates infrastructure review to devops-engineer agent (background)
6. Returns with complete deployment setup plan
Output Format
When activated, I provide:
Immediate Response
- Platform Recommendation: Best fit for project size/budget
- Quick Win: Fastest path to deployment
- Cost Estimate: Monthly hosting costs (indie budget)
Delegated Analysis (Background)
- Launch appropriate agent for infrastructure review
- Provide progress updates
- Synthesize recommendations
Final Output
- Deployment strategy with rationale
- Step-by-step setup (numbered, concrete)
- GitHub Actions workflow (YAML example)
- Environment variables checklist
- Next steps (what to deploy first)
Platform Decision Matrix
| Need | Recommended Platform | Monthly Cost |
|---|
| Static site | GitHub Pages / Netlify | Free |
| Next.js/React | Vercel | Free tier → $20/mo |
| Full-stack app | Render / Railway | $7-20/mo |
| Python/Django | Fly.io / Render | $5-15/mo |
| Database | Supabase / PlanetScale | Free tier → $25/mo |
| Background jobs | Render Cron / Railway | Included |
CI/CD Templates
I'll provide templates for:
GitHub Actions Workflow
name: CI/CD
on: [push, pull_request]
jobs:
test:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Run tests
run: npm test
deploy:
needs: test
if: github.ref == 'refs/heads/main'
runs-on: ubuntu-latest
steps:
- name: Deploy to Vercel
Docker Multi-Stage Build
# Build stage
FROM node:18-alpine AS builder
WORKDIR /app
COPY package*.json ./
RUN npm ci
COPY . .
RUN npm run build
# Production stage
FROM node:18-alpine
WORKDIR /app
COPY --from=builder /app/dist ./dist
COPY package*.json ./
RUN npm ci --production
CMD ["node", "dist/index.js"]
Integration with Existing Workflow
- Auto-activates during
/brainstorm for deployment discussions
- Works with
/next (suggests DevOps implementation steps)
- Integrates with
/done (captures deployment decisions, configs used)
Cost Optimization Tips
Free Tier Strategy:
- Frontend: Vercel/Netlify free tier (hobby projects)
- Backend: Render free tier (spins down after 15 min idle)
- Database: Supabase free tier (500MB, unlimited API requests)
- Monitoring: Sentry free tier (5K events/month)
Total: $0/month for MVP, scales to ~$50/month at 1K users
Remember: This skill auto-activates based on keywords. Keep recommendations indie-friendly (managed platforms, reasonable costs, minimal complexity). Ship fast, scale later.