Compose and deploy full-stack AWS applications in minutes, not days.
IPA is a library of AI-driven skills that automate AWS infrastructure deployment. You describe what you want — a serverless web app, a queue worker, a CI/CD pipeline — and IPA composes the infrastructure, generates the Makefiles, and deploys everything to your account. Choose CloudFormation or Terraform as your IaC engine at init time — the workflow and generated artifacts remain the same.
The output is designed for your path to production. IPA generates standard IaC templates (CloudFormation YAML or Terraform HCL), plain GNU Makefiles with inline CLI calls, and commented .env files — no proprietary abstractions, no runtime dependencies on IPA or a coding agent. An AWS engineer can open any generated file, understand it without IPA context, and integrate it into whatever CI/CD pipeline, change management process, or deployment workflow their organization requires. IPA gets you to a working POC fast; the artifacts it produces are the starting point for production, not a dead end.
⚠️ Important: These samples are not production-ready. They pass automated security scanning at the time of contribution but are not guaranteed to receive ongoing security patches or dependency updates. You must thoroughly review any sample code before deploying to production. See SECURITY.md for details.
A full-stack composition deploys a complete serverless web application in a single session:
- Frontend — React SPA on S3 + CloudFront with OAC
- Backend — FastAPI on Lambda + API Gateway v2 with JWT auth
- Data — DynamoDB tables (feature-flagged)
- Auth — Cognito User Pool with OAuth 2.0 Hosted UI
- Container Registry — ECR for Lambda container images
- Logging — CloudWatch log groups with 30-day retention
Additional stacks layer on top: the queue stack adds a background worker, and stacks compose without conflict.
Four skills, run in sequence:
/ipa-init → /ipa-compose → /ipa-prepare → /ipa-deploy
| Skill | What It Does |
|---|---|
/ipa-init |
Configures the project — writes .env with namespace, environment, region, AWS account, and IaC engine (CloudFormation or Terraform) |
/ipa-compose |
Reads stack skills and generates Makefiles for build, deploy, and teardown. On first run, prompts for security configuration (IAM roles + log bucket). |
/ipa-prepare |
Deploys one-time prerequisite stacks (Cognito, ECR) |
/ipa-deploy |
Builds container images, deploys all stacks, and runs post-deploy wiring |
Not sure what to run next? Use /ipa-help — it inspects your project state and recommends the next skill. Every skill is idempotent — safe to re-run at any time.
- Installation — Install prerequisites (Python 3.12, Node.js, AWS CLI, Claude Code)
- Quickstart — Go from zero to a deployed application in one session
- Concepts — What IPA is, how it is organized, and the lifecycle from inception to production handoff
- Design Principles — The constraints and trade-offs behind IPA's design
- Production-compatible output — every generated artifact (CloudFormation or Terraform, Makefiles,
.env) uses standard tooling and is ready to adopt into Jenkins, GitLab CI, CodePipeline, or any other deployment pipeline without modification - Self-documenting artifacts — inline comments, CloudFormation
Descriptionfields, and Makefile target annotations mean an engineer can read any file and understand the system without consulting IPA documentation - No lock-in — IPA is a composition tool, not a runtime. After deployment, nothing depends on IPA or Claude Code. Delete the skills directory and the system keeps running
- Convention over configuration — sensible defaults for region, naming, and structure mean a builder can run skills without reading a setup guide
- Security by default — least-privilege IAM, encryption at rest, no public-by-default resources
- Evolvable POC — IPA accelerates inception-to-POC; production hardening (HA, multi-AZ, advanced monitoring) is the implementor's next step, shaped by their own requirements