Building a custom agent tool: from schema to deployment

A minimal example (stdio MCP server)

import { Server } from '@modelcontextprotocol/sdk/server'

const server = new Server({ name: 'word-count', version: '0.1.0' })

server.setRequestHandler('tools/list', () => ({
  tools: [{
    name: 'word_count',
    description: 'Count words in a string',
    inputSchema: {
      type: 'object',
      properties: { text: { type: 'string' } },
      required: ['text'],
    },
  }],
}))

server.setRequestHandler('tools/call', async (req) => {
  const text = req.params.arguments.text as string
  return { content: [{ type: 'text', text: String(text.split(/\s+/).length) }] }
})

server.connect(/* stdio */)

Production tips

• Be specific in `description` — it drives model routing
• Prefer many single-purpose tools to one giant Swiss-army tool
• Make side effects idempotent (agents retry)
• Add metrics per tool — see which fire often and which never do