Agent Selection Guide: CI/CD & Production Deployment

Agent Selection Guide: CI/CD & Production Deployment

Use Case: CI/CD pipeline implementation and production deployment for CODITECT Complexity: High - Multi-agent orchestration required

---

🎯 Executive Summary

Task: Implement complete CI/CD pipeline and production deployment infrastructure for CODITECT-CORE.

Recommended Agent Workflow

orchestrator (coordination)
  ↓
devops-engineer (CI/CD pipeline design)
  ↓
cloud-architect (infrastructure & deployment strategy)
  ↓
docker-specialist (containerization - if available) OR devops-engineer
  ↓
codi-documentation-writer (deployment documentation)
  ↓
security-specialist (security review)

Estimated Timeline: 2-3 days for complete implementation Agent Invocations: 5-7 agents sequentially

---

📋 Agent Breakdown by Phase

Phase 1: Planning & Architecture (orchestrator)

Agent: orchestrator Purpose: Coordinate multi-agent workflow and create implementation plan

Invocation

Task(
    subagent_type="orchestrator",
    description="Plan CI/CD deployment workflow",
    prompt="""Create comprehensive implementation plan for CODITECT-CORE CI/CD and production deployment.

Requirements:
- CI/CD pipeline (GitHub Actions)
- Local Docker development environment (Ubuntu LTS + XFCE)
- Production deployment strategy
- Multi-agent task breakdown
- Timeline and dependencies

Deliverable: project-plan.md with task assignments for each agent"""
)

Expected Output

• Complete project plan with phases
• Agent assignments for each task
• Dependency graph
• Timeline estimates
• Risk assessment

---

Phase 2: CI/CD Pipeline Design (devops-engineer)

Agent: devops-engineer or codi-devops-engineer Purpose: Design and implement CI/CD pipelines

Invocation

Task(
    subagent_type="devops-engineer",
    description="Design CI/CD pipeline",
    prompt="""Design comprehensive CI/CD pipeline for CODITECT-CORE using GitHub Actions.

Requirements:
1. Automated testing on push/PR
2. Docker image build and publish
3. Security scanning (containers, dependencies)
4. Deployment to staging/production
5. Rollback capabilities
6. Monitoring integration

Tech Stack:
- GitHub Actions
- Docker Hub or GHCR
- pytest for Python testing
- Security: Trivy, Snyk, or similar

Deliverables:
- .github/workflows/ci.yml
- .github/workflows/deploy.yml
- Dockerfile optimization
- Documentation: CI-CD-PIPELINE.md"""
)

Expected Output

• GitHub Actions workflow files
• Docker build optimization
• Test automation configuration
• Deployment pipeline
• Documentation

---

Phase 3: Infrastructure & Deployment (cloud-architect)

Agent: cloud-architect Purpose: Design production infrastructure and deployment strategy

Invocation

Task(
    subagent_type="cloud-architect",
    description="Design production infrastructure",
    prompt="""Design production deployment infrastructure for CODITECT-CORE.

Requirements:
1. Cloud platform selection (GCP/AWS/Azure)
2. Container orchestration (Docker Compose, K8s, or Cloud Run)
3. Scalability strategy (auto-scaling, load balancing)
4. Database setup (if needed)
5. Monitoring & logging (Prometheus, Grafana, CloudWatch)
6. Disaster recovery & backup
7. Cost optimization

Deliverables:
- Infrastructure as Code (Terraform or equivalent)
- Deployment architecture diagram (C4 model)
- Scaling strategy document
- Cost analysis
- Documentation: PRODUCTION-DEPLOYMENT.md"""
)

Expected Output

• IaC configuration files
• Architecture diagrams
• Deployment strategy
• Cost estimates
• Operations runbook

---

Phase 4: Docker Development Environment (devops-engineer)

Agent: devops-engineer Purpose: Create local Docker development environment

Invocation

Task(
    subagent_type="devops-engineer",
    description="Create Docker dev environment",
    prompt="""Create comprehensive Docker development environment for CODITECT-CORE.

Specifications:
Base Image: Ubuntu 22.04 LTS (Jammy Jellyfish)
Desktop: XFCE (lightweight)
Shell: zsh with oh-my-zsh (jonathan theme)

Required Tools:
- Python 3.10+ with pip, venv
- Node.js 18+ LTS with npm
- Git 2.40+
- Claude Code (npm version: @anthropic/claude-code)
- Docker CLI (for Docker-in-Docker)
- Essential dev tools (curl, wget, vim, nano)
- CODITECT framework pre-installed

User Experience:
- Non-root user 'developer' with sudo
- Pre-configured zsh with jonathan theme
- CODITECT components auto-activated
- Ready-to-use development environment
- Persistent volumes for work

Deliverables:
- Dockerfile
- docker-compose.yml
- entrypoint.sh script
- Documentation: LOCAL-DEVELOPMENT-DOCKER.md
- Quick start script: ./scripts/start-dev-container.sh"""
)

Expected Output

• Multi-stage Dockerfile
• Docker Compose configuration
• Startup scripts
• User-friendly documentation
• Volume management

---

Phase 5: Security Review (security-specialist)

Agent: security-specialist Purpose: Security audit of containers and deployment

Invocation

Task(
    subagent_type="security-specialist",
    description="Security audit deployment",
    prompt="""Perform comprehensive security audit of CODITECT-CORE deployment infrastructure.

Audit Scope:
1. Docker image security (base image, vulnerabilities)
2. CI/CD pipeline security (secrets management, access control)
3. Production deployment security (network, authentication)
4. Dependency vulnerabilities (Python, npm packages)
5. Container hardening (non-root user, minimal attack surface)
6. Secrets management (environment variables, vault)

Deliverables:
- Security audit report
- Vulnerability remediation plan
- Security best practices checklist
- Documentation: SECURITY-DEPLOYMENT.md"""
)

Expected Output

• Security audit report
• Remediation recommendations
• Hardening checklist
• Compliance validation

---

Phase 6: Documentation (codi-documentation-writer)

Agent: codi-documentation-writer Purpose: Create comprehensive deployment documentation

Invocation

Task(
    subagent_type="codi-documentation-writer",
    description="Create deployment documentation",
    prompt="""Create complete deployment documentation for CODITECT-CORE.

Documentation Sections:
1. LOCAL-DEVELOPMENT-DOCKER.md
   - Quick start (5 minutes)
   - Prerequisites
   - Installation steps
   - Configuration
   - Troubleshooting

2. CI-CD-PIPELINE.md
   - Pipeline overview
   - Workflow triggers
   - Build process
   - Testing automation
   - Deployment process

3. PRODUCTION-DEPLOYMENT.md
   - Infrastructure overview
   - Deployment steps
   - Rollback procedures
   - Monitoring setup
   - Operations runbook

4. DEPLOYMENT-CHECKLIST.md
   - Pre-deployment validation
   - Deployment steps
   - Post-deployment validation
   - Health checks

Quality Standards:
- Clear step-by-step instructions
- Code examples for all commands
- Screenshots/diagrams where helpful
- Troubleshooting sections
- Security considerations"""
)

Expected Output

• 4 comprehensive documentation files
• Deployment checklists
• Troubleshooting guides
• Operations procedures

---

🔄 Complete Agent Invocation Sequence

Sequential Execution (Recommended)

# 1. PLANNING - Orchestrator coordinates overall workflow
Task(
    subagent_type="orchestrator",
    description="Plan deployment workflow",
    prompt="Create implementation plan for CI/CD and deployment with agent assignments"
)

# Wait for plan, then proceed with phases:

# 2. CI/CD PIPELINE - DevOps engineer
Task(
    subagent_type="devops-engineer",
    description="Implement CI/CD pipeline",
    prompt="Create GitHub Actions workflows for testing, building, and deployment"
)

# 3. INFRASTRUCTURE - Cloud architect
Task(
    subagent_type="cloud-architect",
    description="Design production infrastructure",
    prompt="Create IaC and deployment architecture for production"
)

# 4. DOCKER DEV ENV - DevOps engineer
Task(
    subagent_type="devops-engineer",
    description="Create Docker development environment",
    prompt="Build Ubuntu XFCE Docker image with CODITECT pre-installed"
)

# 5. SECURITY AUDIT - Security specialist
Task(
    subagent_type="security-specialist",
    description="Security audit",
    prompt="Audit deployment security and provide hardening recommendations"
)

# 6. DOCUMENTATION - Documentation writer
Task(
    subagent_type="codi-documentation-writer",
    description="Create deployment docs",
    prompt="Write comprehensive deployment documentation with examples"
)

---

🛠️ Alternative: Parallel Execution

Some tasks can run in parallel

# Phase 1: Planning (must complete first)
orchestrator_task = Task(subagent_type="orchestrator", ...)

# Phase 2: Parallel execution (after planning)
# Run these simultaneously:
ci_cd_task = Task(subagent_type="devops-engineer", prompt="CI/CD pipeline...")
infra_task = Task(subagent_type="cloud-architect", prompt="Infrastructure...")
docker_task = Task(subagent_type="devops-engineer", prompt="Docker dev env...")

# Phase 3: Sequential (after Phase 2 completes)
security_task = Task(subagent_type="security-specialist", ...)
docs_task = Task(subagent_type="codi-documentation-writer", ...)

Parallel execution saves time but requires coordination

---

📊 Agent Capabilities Matrix

Agent	CI/CD	Docker	Infrastructure	Security	Docs
orchestrator	Plan	Plan	Plan	Plan	Plan
devops-engineer	✅✅✅	✅✅✅	✅✅	✅	-
codi-devops-engineer	✅✅✅	✅✅✅	✅✅	✅	-
cloud-architect	✅	✅	✅✅✅	✅	-
security-specialist	✅	✅	✅	✅✅✅	-
codi-documentation-writer	-	-	-	-	✅✅✅

Legend: ✅✅✅ Expert | ✅✅ Proficient | ✅ Basic | - Not applicable

---

🎯 Expected Deliverables

Code & Configuration

.github/workflows/
├── ci.yml                          # CI pipeline (devops-engineer)
├── deploy-staging.yml              # Staging deployment
└── deploy-production.yml           # Production deployment

distribution/docker/
├── Dockerfile                      # Development container (devops-engineer)
├── docker-compose.yml              # Local dev setup
└── entrypoint.sh                   # Container startup script

infrastructure/
├── terraform/                      # IaC (cloud-architect)
│   ├── main.tf
│   ├── variables.tf
│   └── outputs.tf
└── kubernetes/                     # K8s manifests (if applicable)
    ├── deployment.yml
    └── service.yml

scripts/
├── start-dev-container.sh          # Quick start for developers
├── deploy-to-staging.sh            # Staging deployment
└── deploy-to-production.sh         # Production deployment

Documentation

docs/05-deployment/
├── AGENT-SELECTION-CI-CD-DEPLOYMENT.md  # This file
├── LOCAL-DEVELOPMENT-DOCKER.md          # Docker dev env guide
├── CI-CD-PIPELINE.md                    # CI/CD documentation
├── PRODUCTION-DEPLOYMENT.md             # Production guide
├── SECURITY-DEPLOYMENT.md               # Security guidelines
├── DEPLOYMENT-CHECKLIST.md              # Pre/post deployment
└── OPERATIONS-RUNBOOK.md                # Day-2 operations

---

🚀 Quick Start for Implementation

Option 1: Full Orchestration (Recommended)

# Use orchestrator to coordinate everything
python3 -c "
from anthropic import Task

result = Task(
    subagent_type='orchestrator',
    description='Complete CI/CD deployment implementation',
    prompt='''
    Implement complete CI/CD and deployment infrastructure for CODITECT-CORE:

    1. GitHub Actions CI/CD pipeline
    2. Docker development environment (Ubuntu + XFCE + zsh)
    3. Production deployment (IaC + architecture)
    4. Security audit and hardening
    5. Complete documentation

    Use multi-agent coordination with:
    - devops-engineer (CI/CD, Docker)
    - cloud-architect (infrastructure)
    - security-specialist (security)
    - codi-documentation-writer (docs)

    Deliver production-ready deployment infrastructure.
    '''
)
print(result)
"

Option 2: Manual Agent Invocation

# Invoke agents one at a time, review output, then proceed to next
# See "Complete Agent Invocation Sequence" section above

---

📋 Pre-Implementation Checklist

Before invoking agents, ensure:

• ✅ GitHub repository created and configured
• ✅ Cloud platform account setup (GCP/AWS/Azure)
• ✅ Docker Hub or GHCR account for image registry
• ✅ Secrets configured (API keys, credentials)
• ✅ Production domain/subdomain reserved
• ✅ Monitoring tools accounts (if external)
• ✅ Budget approved for cloud resources
• ✅ Team members have required access

---

🎓 Best Practices

Agent Coordination

1. Start with orchestrator - Let it create the overall plan
2. Review each output - Don't blindly proceed to next agent
3. Iterate if needed - Re-invoke agents for refinements
4. Validate deliverables - Test each component before moving forward
5. Document decisions - Keep track of why certain choices were made

Multi-Agent Tips

• Use orchestrator for task breakdown and coordination
• Invoke agents sequentially for dependent tasks
• Use parallel execution only for independent tasks
• Validate security at each step (not just at the end)
• Keep documentation updated as you go

Quality Gates

After each agent:

• ✅ Review code quality
• ✅ Test functionality
• ✅ Validate security
• ✅ Check documentation completeness
• ✅ Verify alignment with requirements

---

📞 Support & Resources

Agent Documentation

• See .coditect/agents/ for detailed agent capabilities
• Read AGENT-INDEX.md for complete agent catalog
• Check 1-2-3-SLASH-COMMAND-quick-start.md for usage patterns

Related Guides

• CLAUDE-4.5-BEST-PRACTICES.md - Multi-agent orchestration
• MULTI-AGENT-ARCHITECTURE-BEST-PRACTICES.md - Coordination patterns
• 1-2-3-QUICK-START-COMPONENT-ACTIVATION.md - Component management

---

Created: 2025-11-29 Status: Ready for use Quality: Production-grade agent coordination guide