Skip to content

README.md

Latest commit

 

History

History
147 lines (105 loc) · 4.42 KB

README.md

File metadata and controls

147 lines (105 loc) · 4.42 KB
Raw

Vercel Edge Proxy Example

A production-ready chat proxy server that securely forwards requests to Travrse AI. Built with Hono and optimized for Vercel, but also works on other Node.js platforms like Railway, Fly.io, and traditional servers.

This example uses shared flow configurations and utilities from the vanilla-agent-proxy package, making it easy to maintain and extend.

Features

  • Secure API key handling - Never exposes your Travrse API key to the browser
  • Multiple flow configurations:
    • Basic conversational assistant
    • Directive-enabled flow (via flow ID)
    • Shopping assistant with JSON actions
  • Stripe checkout integration - Uses Stripe REST API (no SDK required)
  • CORS support - Configurable allowed origins
  • Auto port selection - Automatically finds available port for local development

Local Development

Prerequisites

  • Node.js 20+ (or use the version specified in .nvmrc if present)
  • pnpm (or npm/yarn)

Setup

  1. Install dependencies:
pnpm install
  1. Create a .env file in this directory:
TRAVRSE_API_KEY=tv_test_xxx
TRAVRSE_FLOW_ID=flow_xxx  # Optional, for directive-enabled flow
STRIPE_SECRET_KEY=sk_test_xxx  # Optional, for checkout functionality
FRONTEND_URL=http://localhost:5173  # Optional, defaults to http://localhost:5173
  1. Start the development server:
pnpm dev

The server will start on port 43111 (or next available port) and automatically reload on file changes.

Available Endpoints

Endpoint Method Description
/api/chat/dispatch POST Basic conversational assistant (default flow)
/api/chat/dispatch-directive POST Directive-enabled flow (requires TRAVRSE_FLOW_ID)
/api/chat/dispatch-action POST Shopping assistant with JSON actions (message, nav_then_click, message_and_click, checkout)
/api/checkout POST Stripe checkout session creation (requires STRIPE_SECRET_KEY)
/form POST Form submission handler for directive demo

Deploying to Vercel

Option 1: Deploy via Vercel CLI

  1. Install Vercel CLI:
npm i -g vercel
  1. Deploy:
vercel
  1. Set environment variables in Vercel dashboard:
    • TRAVRSE_API_KEY (required)
    • TRAVRSE_FLOW_ID (optional)
    • STRIPE_SECRET_KEY (optional)
    • FRONTEND_URL (optional)

Option 2: Deploy via Git Integration

  1. Push your code to GitHub/GitLab/Bitbucket
  2. Import the project in Vercel dashboard
  3. Set the root directory to examples/vercel-edge
  4. Configure environment variables
  5. Deploy

Environment Variables

Set these in the Vercel dashboard (Settings → Environment Variables):

  • TRAVRSE_API_KEY (required): Your Travrse API key
  • TRAVRSE_FLOW_ID (optional): Flow ID for directive-enabled endpoint
  • STRIPE_SECRET_KEY (optional): Stripe secret key for checkout functionality
  • FRONTEND_URL (optional): Your frontend URL for checkout redirect URLs

Deploying to Other Platforms

This example works on any Node.js hosting platform:

Railway

  1. Install Railway CLI: npm i -g @railway/cli
  2. Run railway init
  3. Set environment variables: railway variables set TRAVRSE_API_KEY=xxx
  4. Deploy: railway up

Fly.io

  1. Install Fly CLI: curl -L https://fly.io/install.sh | sh
  2. Run fly launch
  3. Set secrets: fly secrets set TRAVRSE_API_KEY=xxx
  4. Deploy: fly deploy

Traditional Node.js Server

  1. Build the project:
pnpm build

  1. Set environment variables on your server

  2. Start the server:

node dist/server.js

Or use a process manager like PM2:

pm2 start dist/server.js --name chat-proxy

Shared Code

This example uses shared flow configurations and utilities from vanilla-agent-proxy:

  • Flow configs: SHOPPING_ASSISTANT_FLOW from vanilla-agent-proxy/flows
  • Stripe helpers: createCheckoutSession from vanilla-agent-proxy/utils

This eliminates code duplication across deployment examples and makes maintenance easier.

Comparison with Other Examples

Example Best For Deployment Dependencies
vercel-edge (this) Quick start, Node.js platforms Vercel, Railway, Fly.io Node.js runtime
cloudflare-workers Edge computing, global scale Cloudflare Workers None (edge runtime)

For production edge deployment with global low latency, see the cloudflare-workers example.