Vercel

Exposes an MCP endpoint on Vercel to run tools and prompts via a client.
  • javascript

0

GitHub Stars

javascript

Language

4 months ago

First Indexed

3 weeks ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

You can run an MCP (Memory/Contexted Protocol) server on Vercel and interact with it through an MCP client to perform tool-driven tasks. This setup lets you deploy a scalable MCP server that handles your tools and prompts and exposes a simple endpoint for client applications to invoke those tools.

How to use

To use the MCP server from a client, point your MCP client to the deployment URL followed by /mcp. For example, if your deployment is at your-unique.vercel.app, the MCP endpoint is https://your-deployment-url.vercel.app/mcp.

Tools you’ve included in the template demonstrate typical MCP capabilities. The template ships with two example tools to help you get started: roll_dice and get_weather. You can replace these with your own tools and configure prompts, inputs, and resources as needed.

How to install

Prerequisites: you need a development environment with Node.js installed. Ensure you have a Vercel account and can deploy projects.

Step 1: Prepare the project

  • Create or clone the MCP server project that targets Vercel. This template is designed for deployment to Vercel and includes an API route for MCP interactions.

Step 2: Configure your server

  • Update the server entry (for example api/server.ts) with your tools, prompts, and resources according to your MCP setup. Use the MCP SDK patterns to register tools and define how they should be executed.

Step 3: Deploy to Vercel

  • Ensure Fluid compute is enabled for efficient execution on Vercel.

Step 4: Adjust deployment time limits (if needed)

  • If you are on a Pro or Enterprise plan, open vercel.json and set the max duration to 800 to accommodate longer-running tasks.

Step 5: Run locally for development

  • Use the local development workflow to test your MCP server with Vercel’s dev tooling:
  • vercel dev

Step 6: Validate the MCP endpoint

  • After deployment, verify that your MCP endpoint is reachable at the URL shown in Step 1 and use the sample client to test invocations.

Additional notes and usage patterns

Local development and testing can be done with the sample client. You can experiment with the included tools to validate input handling, prompts, and tool results before adding your own tools.

Sample client usage is provided to try invocations against your deployed MCP server. The client demonstrates a typical invocation pattern against the /mcp endpoint.

Example client invocation

node scripts/test-client.mjs https://mcp-on-vercel.vercel.app

Available tools

roll_dice

Rolls an N-sided die (minimum 2 sides) to produce a random result for a given number of faces.

get_weather

Fetches current weather data for a location using coordinates or city name, integrating with a weather API.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational