blissful-infra deploy

blissful-infra deploy takes the app you built locally and ships it to a real cloud environment. Set deploy.target in your blissful-infra.yaml to choose the platform, then run the same command every time.

Usage

blissful-infra deploy [name] [options]

[name] is optional. If omitted, the CLI reads blissful-infra.yaml from the current directory.

Options

Flag	Description
--dry-run	Print what would be deployed without making any changes

Configuration

Set deploy.target in blissful-infra.yaml before running deploy:

name: my-app
backend: spring-boot
frontend: react-vite
database: postgres
deploy:
  target: cloudflare   # cloudflare | vercel | aws

You can set the target at scaffold time so the project is ready from the start:

blissful-infra start my-app --deploy-target cloudflare

Deploy targets

Cloudflare

Prerequisites: wrangler CLI installed and authenticated.

npm install -g wrangler
wrangler login

What deploys:

Local	Cloudflare
React + Vite frontend	Cloudflare Pages
Express / Hono backend	Cloudflare Worker
Postgres database	Cloudflare D1 (SQLite)
Redis cache	Cloudflare KV

The CLI calls wrangler for each step, you don’t need to know the wrangler commands yourself. A wrangler.toml is generated in frontend/ and backend/ at scaffold time if you used --deploy-target cloudflare. If you’re deploying an existing project, the CLI generates them on first deploy.

Config block (optional):

deploy:
  target: cloudflare
  cloudflare:
    accountId: your-cf-account-id
    workerName: my-app-api
    pagesProject: my-app-frontend

If accountId is omitted, wrangler uses your default account from wrangler login.

Vercel

Prerequisites: vercel CLI installed and authenticated.

npm install -g vercel
vercel login

What deploys: frontend via vercel build + vercel deploy --prebuilt --prod, backend via vercel deploy --prod.

Vercel uses real Postgres (not SQLite), so no DDL translation is needed. Configure DATABASE_URL, Redis (Upstash), and queue (QStash) environment variables in your Vercel project dashboard.

Config block (optional):

deploy:
  target: vercel
  vercel:
    orgId: your-org-id
    projectId: your-project-id

AWS

Prerequisites: AWS CLI and CDK installed and configured.

brew install awscli
npm install -g aws-cdk
aws configure

What deploys: CDK stacks via cdk deploy --all. The scaffold generates CDK stacks for ECS Fargate (backend), S3 + CloudFront (frontend), and RDS Postgres (database).

Config block (optional):

deploy:
  target: aws
  aws:
    region: us-east-1
    cluster: my-app-cluster

Examples

# Deploy from the project directory
cd my-app
blissful-infra deploy

# Deploy by project name from parent directory
blissful-infra deploy my-app

# Preview without making any changes
blissful-infra deploy --dry-run

Module portability

The local modules map to platform-native equivalents so your application code doesn’t change when you switch targets:

Module	Local	Cloudflare	Vercel	AWS
Frontend	nginx	CF Pages	Vercel	S3 + CloudFront
Backend	Docker	CF Worker	Vercel Functions	ECS Fargate
Database	Postgres	D1 (SQLite)	Vercel Postgres	RDS
Cache	Redis	CF KV	Upstash Redis	ElastiCache
Queue	Kafka	CF Queues	Upstash QStash	SQS

To switch platforms, change deploy.target in blissful-infra.yaml and run blissful-infra deploy again.

Error handling

Missing target: If deploy.target is local-only or not set, the CLI prints the config block you need to add and exits.

Missing prerequisite: If wrangler, vercel, or aws is not installed, the CLI prints the exact install command and exits.

Deploy failure: The CLI surfaces the underlying error output and exits with the same code as the failing tool.