Deployment Orchestrator Skill

Activation Method

This skill activates when the hook system injects the codeword:

[ACTIVATE:DEPLOYMENT_ORCHESTRATOR_V1]

This occurs when:

• Phase 10 E2E tests pass
• Task #26 (deployment) is active
• Automatically triggered after Phase 5 completion

Worktree Isolation Requirements

CRITICAL: This skill MUST operate in a dedicated worktree

phase-11-task-1

:

# Before skill activation:
./lib/worktree-manager.sh create 6 1
cd ./worktrees/phase-11-task-1

# Validate isolation:
./hooks/worktree-enforcer.sh enforce

# Deployment orchestration with isolation

Deployment Isolation Strategy

1. Secure deployment environment: Deployment operations in isolated workspace
2. Artifact isolation: Build and deployment artifacts contained within worktree
3. Configuration isolation: Deployment configs managed without contamination
4. Rollback preparation: Rollback scripts and artifacts isolated and ready
5. Production merge: Final deployment merged to main with full validation

What This Skill Does

Automates Phase 6: Production deployment in isolated worktree (fully autonomous)

• Infrastructure validation (Docker, Kubernetes, services)
• Container build and startup (docker-compose up)
• Health check validation (all services healthy)
• Staging deployment (validate) in isolated environment
• Canary deployment (monitor) with isolated deployment artifacts
• Production rollout (gradual or immediate) from clean workspace
• Automatic progression through all deployment stages
• Rollback capability (if issues detected) with isolated rollback scripts
• NEW: Isolated deployment environment prevents contamination
• NEW: Secure artifact management within worktree boundaries

Execution Flow

Stage 0: Automated Validation
         - Run security validation
         - Run load testing
         - Validate performance targets
         - BLOCK if any validation fails
Stage 1: Infrastructure Setup
         - Build Docker containers (docker-compose build)
         - Start all services (docker-compose up -d)
         - Validate health checks
         - Verify connectivity
Stage 2: Pre-Deployment Validation
         - Check all tests passing
         - Validate production readiness score
Stage 3: Staging Deployment
         - Deploy to staging environment
         - Run smoke tests
         - Validate monitoring
         - Auto-proceed on success
Stage 4: Canary Deployment
         - Deploy to 5% of production traffic
         - Monitor metrics
         - Compare metrics vs baseline
         - Auto-proceed if metrics healthy
Stage 5: Production Rollout
         - Gradual rollout (10% → 50% → 100%)
         - OR immediate (100%)
         - Monitor continuously
Stage 6: Post-Deployment Validation
         - Verify all services healthy
         - Confirm metrics normal
         - Generate completion report

Deployment Strategy

Stage 0: Automated Validation (MUST RUN FIRST)

# Run all automated validators - BLOCKS deployment if any fail
echo "==============================================================================="
echo "Stage 0: Automated Validation"
echo "==============================================================================="

# Security validation
echo ""
echo "Running security validation..."
./hooks/security-validator.sh || {
    echo "❌ Security validation FAILED - deployment blocked"
    exit 1
}

# Load testing
echo ""
echo "Running load tests..."
./hooks/load-test-validator.sh || {
    echo "❌ Load testing FAILED - deployment blocked"
    exit 1
}

# Performance validation
echo ""
echo "Validating performance targets..."
./hooks/performance-validator.sh || {
    echo "⚠️  Performance validation failed but continuing (check PRD requirements)"
}

echo ""
echo "✅ All automated validations PASSED"
echo "Proceeding to infrastructure setup..."

CRITICAL:

• These validators MUST pass before proceeding
• Security failures = hard block
• Load test failures = hard block
• Performance failures = warning (may continue if targets not in PRD)

---

Infrastructure Setup

# Build and start Docker containers
docker-compose build
docker-compose up -d

# Wait for services to be healthy
timeout 300 bash -c 'until docker-compose ps | grep -v "unhealthy\|starting"; do sleep 10; done'

# Verify all services running
docker-compose ps
docker-compose logs --tail=50

Staging

# Deploy to staging
./scripts/deploy.sh staging

# Run smoke tests
npm test:smoke

# Validate
./scripts/health-check.sh staging

Canary (automatic)

# Deploy 5% traffic
./scripts/deploy.sh canary --traffic=5

# Monitor metrics automatically
# Watch: error rate, latency, throughput

# Auto-proceed if metrics healthy
# Auto-rollback if metrics degrade

Production (automatic)

# Gradual rollout
./scripts/deploy.sh prod --traffic=10
# Monitor, auto-proceed if healthy
./scripts/deploy.sh prod --traffic=50
# Monitor, auto-proceed if healthy
./scripts/deploy.sh prod --traffic=100

# OR immediate (if canary validated)
./scripts/deploy.sh prod --traffic=100

Automatic Validation Gates

Gate 1: Staging → Canary (Auto)

Required checks (automated):

• ✅ Staging deployment successful
• ✅ Smoke tests passing
• ✅ No errors in logs
• ✅ Monitoring dashboards healthy

Action: Auto-proceed to canary if all checks pass

Gate 2: Canary → Production (Auto)

Required checks (automated):

• ✅ Canary metrics stable
• ✅ Error rate ≤ baseline
• ✅ Latency ≤ baseline +10%

Action: Auto-proceed to production if metrics healthy

Rollback Triggers

Automatic rollback if:

• Error rate > baseline + 50%
• Latency > baseline + 100%
• Critical service down > 1 min

Manual rollback:

./scripts/rollback.sh

Monitoring Dashboard

Key metrics:

• Request rate (req/sec)
• Error rate (%)
• Latency (p50, p95, p99)
• CPU usage (%)
• Memory usage (%)
• Database connections

Time Estimates

Phase	Duration
Staging	30 min
Canary	24 hours
Production	2-8 hours
Total	25-32 hours

Completion Signal

{
  "phase": 6,
  "status": "success",
  "summary": {
    "deployed_to": "production",
    "traffic": 100,
    "health": "green",
    "rollbacks": 0
  },
  "pipeline_complete": true
}

Output Files

.taskmaster/
├── DEPLOYMENT_REPORT.md
└── .signals/phase11-complete.json

logs/
├── deployment-staging.log
├── deployment-canary.log
└── deployment-production.log

CRITICAL: Fully Autonomous Deployment

DO NOT ASK THE USER FOR PERMISSION AT ANY DEPLOYMENT STAGE.

This skill automatically progresses through all deployment stages:

1. Staging → (auto-proceed on success)
2. Canary → (auto-proceed if metrics healthy)
3. Production → (auto-proceed with gradual rollout)

When deployment is complete, output:

✅ PHASE 6 COMPLETE - DEPLOYMENT SUCCESSFUL
[SIGNAL:PHASE11_COMPLETE]
🎉 PIPELINE COMPLETE - Application deployed to production

The entire pipeline from Phase 2 onward is fully autonomous. Do not ask "Would you like to proceed?" at any stage - just proceed automatically.

See Also

• Pipeline Orchestrator (triggers this, manages approvals)
• E2E Validator (Phase 5, provides GO decision)