Chatgpt App:deploy

Deploy ChatGPT App

You are helping the user deploy their ChatGPT App to Render.

Prerequisites

Before deploying:

1. Validation passed (/chatgpt-app:validate)
2. Tests passed (/chatgpt-app:test)
3. All changes committed to git
4. Environment variables ready

Workflow

1. Pre-Flight Check

## Deployment Pre-Flight

- [ ] Validation: PASS/FAIL
- [ ] Tests: PASS/FAIL
- [ ] Git status: clean
- [ ] Migrations: ready

2. Generate Render Configuration

Create render.yaml:

services:
  - type: web
    name: {app-name}
    runtime: docker
    healthCheckPath: /health
    envVars:
      - key: DATABASE_URL
        fromDatabase:
          name: {app-name}-db
          property: connectionString

databases:
  - name: {app-name}-db
    plan: starter

3. Generate Dockerfile

Create production Docker image.

4. Deploy

Option A: Render MCP (Automated) If user has Render MCP:

1. Set workspace
2. Create database
3. Create web service
4. Set environment variables

Option B: Render Dashboard (Manual) Guide through web UI.

5. Verify Deployment

1. Health check: curl https://{app}.onrender.com/health
2. MCP endpoint: curl https://{app}.onrender.com/mcp
3. Tool discovery via MCP Inspector

6. Configure ChatGPT Connector

1. Go to ChatGPT Settings → Connectors
2. Enable Developer Mode
3. Add connector: https://{app}.onrender.com/mcp
4. Test with golden prompts

Environment Variables

Required:

• NODE_ENV=production
• PORT=8787
• DATABASE_URL (from Render)

Auth (if enabled):

• Auth0 or Supabase credentials

Update State

Save deployment info to .chatgpt-app/state.json:

{
  "deployment": {
    "platform": "render",
    "status": "deployed",
    "mcpEndpoint": "https://{app}.onrender.com/mcp"
  }
}