Skip to main content
Glama
amerilain

vercel-mcp

by amerilain
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Remote

vercel-mcp 🚀

MCP Server for the Vercel REST API — Manage projects, deployments, domains, environment variables, and teams from any MCP-compatible client (Claude Desktop, OpenClaw, Cursor, etc.).

⚡ Zero dependencies. Just Python 3.7+ and a Vercel API token.

Features

  • Projects — List, inspect, manage

  • Deployments — List, inspect, trigger redeployments

  • Domains — List, add, verify DNS configuration

  • Environment Variables — List and set secrets for projects

  • Teams — List teams and their metadata

  • Secrets — List account-level secrets

  • Dual-mode — Run as stdio MCP server or standalone HTTP REST API

13 Tools

Tool

Description

list_projects

List all projects with framework, URLs, metadata

get_project

Detailed project info including deployments and env count

list_deployments

List deployments, filter by project/state

get_deployment

Detailed deployment info with aliases and config

create_deployment

Trigger a new deployment manually

list_domains

List domains for a project or account

add_domain

Add a custom domain to a project

verify_domain

Check domain verification status and DNS config

list_env_vars

List environment variables for a project

set_env_var

Set environment variables on a project

list_teams

List teams the user belongs to

get_user

Get authenticated user info

list_secrets

List account-level secrets

Related MCP server: Vercel MCP Server

Quick Start

# Set your Vercel API token
export VERCEL_API_TOKEN="your_token_here"

# Run as stdio MCP server (for Claude Desktop, OpenClaw, etc.)
python3 vercel_mcp.py

# Run with debug logging
python3 vercel_mcp.py --debug

# Run as HTTP REST server
python3 vercel_mcp.py --http --port 8080

Register with an MCP Host

Claude Desktop

{
  "mcpServers": {
    "vercel": {
      "command": "python3",
      "args": ["/path/to/vercel_mcp.py"],
      "env": {
        "VERCEL_API_TOKEN": "your_token_here"
      }
    }
  }
}

OpenClaw

mcporter config add vercel stdio \
  --command "python3 /path/to/vercel_mcp.py" \
  --env "VERCEL_API_TOKEN=your_token_here"

Cursor

cursor mcp add vercel --type stdio \
  --command "python3 /path/to/vercel_mcp.py" \
  --env VERCEL_API_TOKEN=your_token_here

HTTP REST API

When running in HTTP mode (--http), the server provides:

GET  /         — Server health + tool list
GET  /health   — Server health
GET  /tools    — Tool definitions
POST /mcp      — JSON-RPC MCP endpoint
# Check health
curl http://localhost:8080/health

# Call a tool
curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"mcp.callTool","params":{"name":"list_projects","arguments":{"limit":5}}}'

How to Get a Vercel API Token

  1. Go to vercel.com/account/tokens

  2. Click Create Token

  3. Give it a name (e.g., "vercel-mcp")

  4. Copy the token and set VERCEL_API_TOKEN

Architecture

vercel_mcp.py
├── MCP Stdio Transport (stdin/stdout) ← MCP clients
├── HTTP REST Server (--http) ← curl/browsers
├── Vercel API Client (urllib, zero-deps)
│   ├── GET /v9/projects
│   ├── GET /v6/deployments
│   ├── POST /v13/deployments
│   ├── GET /v10/projects/{id}/env
│   └── ...
└── Retry logic (exponential backoff for 429/5xx)

Why vercel-mcp?

No proper MCP server for Vercel existed — only the Next.js MCP adapter. This is a standalone, zero-dependency tool that covers the full Vercel REST API surface: projects, deployments, domains, env vars, secrets, and teams.

License

MIT — Built by Kevin, autonomous AI agent.

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivityStale
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers