---

name: dev-swarm-stage-deployment description: Deploy to production, configure infrastructure, set up monitoring, and prepare for launch. Use when starting stage 11 (deployment) or when user asks to deploy, go-live, or launch the product.

Stage 11 - Deployment

Deploy the fully developed and tested product to production environment, configure production infrastructure, set up monitoring and alerting, and prepare for launch including documentation and marketing readiness.

When to Use This Skill

• User asks to start stage 11 (deployment)
• User wants to deploy to production or go-live
• User asks about launch preparation or marketing readiness
• User wants to set up production monitoring or infrastructure

Your Roles in This Skill

See dev-swarm/docs/general-dev-stage-rule.md for role selection guidance.

Role Communication

See dev-swarm/docs/general-dev-stage-rule.md for the required role announcement format.

Pre-Stage Check

Before starting, verify previous stages:

1. Check if stages 00-init-ideas/ through 10-sprints/ have content (not just .gitkeep)
2. If any previous stage is empty and has no SKIP.md:
   • Ask user: "Stage {XX} is not complete. Would you like to skip it or start from that stage first?"

Instructions

Step 1: Context Review

Read all files to understand the project:

• ideas.md
• 00-init-ideas/*.md through 10-sprints/*.md - All markdown files
• Critical: Review 09-devops/execution-plan.md to identify any Part 2 items (deferred to Stage 11) that need to be executed before deployment
• Source Code Location: Use {SRC}/ as the default source code root if not specified in project documentation

Step 2: Create Stage Proposal

General Rules: See dev-swarm/docs/general-dev-stage-rule.md → "Create Stage Proposal Rules" section.

If this stage is skipped (has SKIP.md), execute the next non-skipped stage's agent skill. Otherwise, create the file 11-deployment/README.md with the following content:

2.1 Stage Goal

Brief the goal in 2-3 paragraphs:

• What this stage aims to achieve (production deployment, go-live readiness)
• Why proper deployment preparation is critical for product success
• How this stage builds upon all previous stages
• What deliverables will be produced (live production system, documentation, launch plan)

2.2 Deferred Infrastructure Review

Check 09-devops/execution-plan.md Part 2 and list items that were deferred:

• Which infrastructure items need to be set up now
• What credentials/access is required
• Estimated costs for production infrastructure

2.3 File Selection

Select files from these options based on project needs:

Deployment Planning:

• deployment-plan.md - Detailed deployment plan including sequence, dependencies, and timeline
• production-config.md - Production environment configuration (environment variables, feature flags)
• deployment-diagram/ - Deployment architecture and flow diagrams (follow dev-swarm/docs/mermaid-diagram-guide.md)
• execution-plan.md - Checklist of deployment tasks with pre-deployment and post-deployment sections

Cloud Infrastructure (Production):

• aws-production-setup.md - AWS production setup (EC2, RDS, S3, CloudFront, etc.)
• runpod-production-setup.md - RunPod GPU production setup (if applicable)
• ssl-certificate-setup.md - SSL/TLS certificate setup and renewal strategy
• domain-dns-setup.md - Domain registration and DNS configuration

Operations & Reliability:

• monitoring-alerting.md - Production monitoring, alerting rules, and on-call procedures
• backup-strategy.md - Backup schedules, retention policies, and disaster recovery plan
• rollback-plan.md - Rollback procedures for failed deployments
• security-hardening.md - Production security measures and hardening checklist
• performance-optimization.md - Performance tuning and optimization guide

Verification & Quality:

• post-deployment-verification.md - Smoke tests, health checks, and verification procedures
• go-live-checklist.md - Final checklist before going live

Documentation:

• user-documentation.md - End-user documentation and help guides
• admin-documentation.md - Admin/ops documentation for system management
• api-documentation.md - Public API documentation (if applicable)

Launch & Marketing:

• marketing-readiness.md - Marketing materials checklist (landing page, screenshots, copy)
• launch-announcement.md - Launch communication plan (social media, press release, email)

Mobile App Store Release (if applicable):

• app-store-release.md - Apple App Store submission process
• google-play-release.md - Google Play Store submission process
• mobile-release-checklist.md - Pre-submission checklist

For each selected file, provide:

• Short description
• Why it's essential for this project
• Key information it should include

Note: Not all files may be necessary. Select based on project complexity and launch requirements.

2.4 Request User Approval

Ask user: "Please check the Stage Proposal in 11-deployment/README.md. Update it directly or tell me how to update it."

Step 3: Execute Stage Plan

Once user approves 11-deployment/README.md:

3.1 Create All Planned Files

Create each file listed in the approved README:

• For .md files: Write comprehensive, well-structured documentation with step-by-step instructions
• For diagram folders: Follow dev-swarm/docs/mermaid-diagram-guide.md to create related diagrams files

Quality Guidelines:

• Base configurations on architecture and tech-specs from previous stages
• Include security best practices for production environment
• Document all environment variables and secrets required
• Provide clear rollback procedures for each deployment step
• Include verification steps after each major deployment action
• Reference sprint deliverables from 10-sprints/
• Ensure monitoring covers all critical system components
• Document estimated costs for production infrastructure
• Source Code Management: Package source code from {SRC}/ directory (default) or as specified in project setup
• Git Management: Create/update .gitignore file to exclude unnecessary files (logs, build artifacts, secrets, node_modules, etc.) from commits

3.2 Request User Approval for Files

After creating all files:

• Provide a summary of what was created
• Highlight key deployment decisions
• List any credentials or access needed for deployment
• Ask: "Please review the deployment documentation. You can update or delete files, or let me know how to modify them."

Step 4: Finalize Documentation

Once user approves all files:

4.1 Documentation Finalization

• Sync 11-deployment/README.md to remove any deleted files
• Ensure all files are complete and well-formatted
• Verify deployment plan aligns with architecture design
• Confirm all setup steps are documented with prerequisites
• Check that all diagrams render correctly

4.2 Announce Documentation Complete

Inform user:

• "Deployment documentation is complete and ready for execution"
• Summary of files created
• List of infrastructure to be provisioned

Step 5: Create Execution Plan

Create 11-deployment/execution-plan.md with deployment tasks organized into these sections:

Part 1: Pre-Deployment (Infrastructure from Stage 09) Items deferred from 09-devops/execution-plan.md Part 2

Part 2: Deployment Execution

• Source code preparation (from {SRC}/ directory by default, ensure .gitignore excludes unnecessary files)
• Code merge, database migration, application deployment, CDN configuration, DNS cutover

Part 3: Post-Deployment Verification

• Health checks, smoke tests, performance baseline, security scan, monitoring verification

Part 4: Go-Live Activities

• Remove maintenance mode, enable traffic, execute launch announcement

Part 5: Mobile App Store Release (if applicable)

• App Store Connect and Google Play metadata, submissions, review monitoring, release

For each item, include:

• Description of what will be done
• Prerequisites (dependencies, credentials, approvals)
• Rollback procedure if it fails
• Success criteria

Ask user: "Please review the deployment execution plan. Confirm which items to execute and provide any required credentials."

Step 6: Execute Deployment

Once user approves, execute in order with checkpoints after each part:

• Execute Part 1 (deferred infrastructure) first
• Execute Part 2 (deployment) with verification
• Execute Part 3 (post-deployment verification) - all must pass before go-live
• Execute Part 4 (go-live activities)
• Execute Part 5 (mobile releases) if applicable

Execution Guidelines:

• Execute steps in order, verify each before proceeding
• Document any issues and resolutions immediately
• Have rollback plan ready at each step
• Maintain communication with user throughout
• Stop and consult user if any critical issues arise

Step 7: Update Documentation

After deployment:

• Record actual resource IDs, URLs, endpoints created
• Update configuration files with real production values (excluding secrets)
• Document any deviations from the original plan
• Create operations runbook with troubleshooting steps
• Update user documentation with production URLs

Step 8: Complete Stage & Project

Once user confirms deployment is successful:

8.1 Announce Completion

Inform user:

• "Stage 11 (Deployment) is complete - Product is LIVE!"
• Production URL and access information
• Summary of infrastructure deployed
• List of documentation delivered
• Recommended post-launch activities
• "Congratulations on your successful launch!"

Stage Completion Rules

See dev-swarm/docs/general-dev-stage-rule.md for stage completion, commit, and skip rules.

Key Principles

• Prioritize safety with multiple checkpoints before and after deployment
• Document everything for future maintenance
• Ensure rollback capability at every step
• Communicate clearly with user throughout the process
• Verify production system is fully operational before announcing completion