DeployIt

Deploy your frontend effortlessly.DeployIt is a static site hosting platform where users authenticate with GitHub, import repositories, build them on AWS ECS Fargate, and serve the generated static artifacts from Amazon S3.

Architecture

graph TB
    subgraph Dashboard
        DASH["Next.js 16 + shadcn UI<br/>:3000"]
    end

    subgraph APILayer["API Layer"]
        APISERVER["api-server..."]
    end

    subgraph Pipeline["Build Pipeline"]
        QUEUE[("Redis<br/>build_queue")]
        ORCH["orchestrator<br/>BRPOP worker<br/>:3003"]
        ECS["ECS Fargate<br/>build-agent task<br/>(or local subprocess)"]
    end

    subgraph Edge["Edge Serving"]
        PROXY["edge-proxy<br/>Express :8000"]
        S3[("S3 or /tmp/")]
    end

    subgraph Data["Data Plane"]
        RDS[("PostgreSQL<br/>(Docker)")]
        REDIS[("Redis<br/>(Docker)")]
    end

    DASH -->|REST + SSE| API
    API -->|Prisma| RDS
    API -->|RPUSH| QUEUE
    ORCH -->|BRPOP| QUEUE
    ORCH -->|RunTask or subprocess| ECS
    ECS -->|git clone| GitHub[("GitHub")]
    ECS -->|upload| S3
    PROXY -->|slug lookup| RDS
    PROXY -->|serve| S3

Loading

Quick Start

1. Install Dependencies & Setup Environment

# Install root dependencies
bun install

# Configure environment variables
cp .env.example .env
# Edit .env and fill in required GitHub OAuth & AWS credentials

# Symlink .env to all individual services
for svc in api-server orchestrator build-agent edge-proxy dashboard; do
  ln -sf ../.env $svc/.env
done

---

Step 2: Run AWS Setup Script

The scripts/setup-aws.sh script creates all necessary AWS resources.

cd /home/rishisulakhe/projects/vercel

# Make sure you're logged in
aws sts get-caller-identity

# Run the setup script
./scripts/setup-aws.sh

What the script creates:

Resource	Purpose	Cost
S3 Bucket	Store build artifacts	~$0.01/GB/month
ECR Repository	Store build-agent Docker image	~$0.10/GB/month
ECS Cluster	Orchestrate build tasks	$0 (Fargate pay-per-task)
IAM Task Execution Role	Allow ECS to pull images	Free
IAM Task Role	Allow build-agent to write to S3	Free
ECS Task Definition	Define build-agent container config	Free
CloudWatch Log Group	Store build logs	~$0.50/GB

---

Step 3: Update .env File

Create/update your .env file:

cp .env.example .env

Edit .env and ensure these are set:

# --- GitHub OAuth ---
GITHUB_CLIENT_ID="Ov23liLuiYXJ3T4h87rX"
GITHUB_CLIENT_SECRET="72c1b2b89c687a1b6ce0c6dcf653d2bd18f927dd"
NEXT_PUBLIC_GITHUB_CLIENT_ID="Ov23liLuiYXJ3T4h87rX"
NEXT_PUBLIC_GITHUB_REDIRECT_URI="http://localhost:3000/api/auth/callback/github"
GITHUB_REDIRECT_URI="http://localhost:3000/api/auth/callback/github"
NEXTAUTH_SECRET="bjeOoKTmIIMEmTGch06kUpDVt++uG5T+4xFs1lV2grw="

# --- AWS (from setup-aws.sh output) ---
AWS_REGION="ap-south-2"
S3_ARTIFACTS_BUCKET="deployit-rishi-artifacts"
ECS_CLUSTER="vercel-clone"
ECS_BUILD_TASK_DEFINITION="vercel-clone-build-agent"
ECS_BUILD_TASK_SUBNETS="subnet-aaa,subnet-bbb"
ECS_BUILD_TASK_SECURITY_GROUPS="sg-ccc"

Important: Keep these as-is for AWS mode:

S3_ARTIFACTS_BUCKET="deployit-rishi-artifacts"  # NOT "local"
ECS_BUILD_TASK_SUBNETS="subnet-xxx"           # NOT empty
EDGE_PROXY_BACKEND_BASE_URL=""                 # empty = serve from S3 directly

---

4. Build & Push Build Agent to Amazon ECR

# Authenticate Docker with ECR
aws ecr get-login-password --region "$AWS_REGION" | \
  docker login --username AWS --password-stdin "$(echo "$ECR_URI" | cut -d/ -f1)"

# Build and push the container image
cd build-agent
docker build -t "$ECR_URI:latest" .
docker push "$ECR_URI:latest"
cd ..

---

5. Start Backing Services & Migrate Database

# Start PostgreSQL, Redis, Prometheus, and Grafana
docker compose up -d

# Generate Prisma client and apply database migrations
cd api-server
bunx prisma generate --schema=prisma/schema.prisma
bunx prisma migrate deploy --schema=prisma/schema.prisma
cd ..

---

6. Start Application Services

Run each command in a separate terminal:

bun run dev:api        # API Server:     http://localhost:3001
bun run dev:orch       # Orchestrator:   http://localhost:3003
bun run dev:proxy      # Edge Proxy:     http://localhost:8000
bun run dev:dashboard  # Dashboard:      http://localhost:3000

---

7. Access the Platform

• Dashboard: http://localhost:3000 (Sign in with GitHub and deploy a repository)
• Deployed Sites: http://localhost:8000/<project-slug>/

Services

Package	Runtime	Port	Role
dashboard/	Node 22	3000	Next.js 16 + shadcn UI, GitHub OAuth, SSE logs
api-server/	Bun	3001	Hono REST API + Prisma
orchestrator/	Bun	3003	Redis BRPOP → ECS/subprocess dispatcher
build-agent/	Bun	—	git clone → build → S3 upload
edge-proxy/	Bun	8000	Subdomain/path routing → S3 filesystem

Documentation

| Document | Description | |---|---|---| | docs/architecture.md | System design, data flow, decision log | | docs/runbooks/local-dev.md | Development setup, troubleshooting | | docs/aws-deployment-guide.md | Complete AWS setup guide | | docs/runbooks/deploy-project.md | Deploying via UI or API | | docs/runbooks/secrets.md | GitHub OAuth setup | | docs/runbooks/troubleshooting.md | Common issues |

Summary

This project provides a Vercel-style deployment workflow with a local application control plane and AWS-powered build infrastructure.

                  LOCAL
┌───────────────────────────────────┐
│ Dashboard                         │
│    ↓                              │
│ API ─────→ PostgreSQL             │
│  │                                │
│  └────→ Redis → Orchestrator ─────┼──── AWS
│                                   │      │
│ Edge Proxy ←───────────────────────┼── S3 │
└───────────────────────────────────┘      │
                                           │
                                    ECS Fargate
                                           │
                                      build-agent
                                           │
                                      GitHub + ECR

License

MIT.