Build a sports research agent with the Vercel AI SDK and MCP

npm i ai @ai-sdk/mcp
MODEL=<your model id> npx tsx ai_sdk_mcp_research.ts

// Slate research with the Vercel AI SDK (v7) and the DeepChamp MCP server. The free tools need no key.
// Set MODEL to any model id your AI SDK provider or the AI Gateway accepts.
import { createMCPClient } from '@ai-sdk/mcp';
import { generateText, isStepCount, type ToolSet } from 'ai';

const system =
  'You write short pre-game research briefings from DeepChamp tool results. Use only what the tools return. Say when the data was current (meta.as_of). Never state odds, picks or outcomes, and never claim a result. End with: Research only, not betting advice. 21+. Help: 1-800-MY-RESET.';

const mcp = await createMCPClient({ transport: { type: 'http', url: 'https://deepchampai.com/mcp' } });

try {
  const tools = (await mcp.tools()) as ToolSet; // MCP tools carry no static input types, so name the set
  const { text } = await generateText({
    model: process.env.MODEL as string,
    system,
    tools,
    stopWhen: isStepCount(6),
    prompt: 'Brief me on today’s NBA slate: which games have rest or schedule angles worth checking first?',
  });
  console.log(text);
} finally {
  await mcp.close();
}