Brave Search MCP Server on Railway

This is a Brave Search MCP HTTP service deployed on Railway.

Deployment Methods

Method 1: Deploy via GitHub (Recommended)

1. Fork or upload this directory to a GitHub repository

2. Log in to the Railway Dashboard

3. Click New Project → Deploy from GitHub repo

4. Select the repository you uploaded

   • Railway will automatically detect package.json and run npm start

   • Railway will automatically set the PORT environment variable

Method 2: Via Railway CLI

npm i -g @railway/cli
railway login
railway init
railway up

Related MCP server: Search MCP

Environment Variables

You must set the following in the Railway Dashboard:

There is no need to manually set PORT; Railway handles it automatically.

Obtaining a Brave API Key

1. Go to Brave Search API to register

2. Select the Search plan (suitable for AI Agents)

3. Generate an API Key in the Developer Dashboard

After Deployment

Railway will provide a domain like https://xxx.up.railway.app.

The MCP endpoint address is: https://xxx.up.railway.app/mcp

Local testing:

curl -s -X POST https://xxx.up.railway.app/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'

Adding MCP to Hermes

After deployment, run:

# 添加 MCP 服务器
hermes mcp add brave-search --url https://你的域名.up.railway.app

# 确认无认证需求（API Key 在服务端设置）

This allows Hermes to perform searches via the Brave Search MCP service on Railway.