MCP Server Examples

Real-world examples of MCP servers on Universal API.

Echo Server (Minimal)

The simplest possible MCP server — great for testing:

const { McpServer } = require("@modelcontextprotocol/sdk/server/mcp.js");
const { z } = require("zod");

function createMcpServer() {
  const server = new McpServer({ name: "echo", version: "1.0.0" });

  server.registerTool("echo", {
    title: "Echo",
    description: "Echo back the provided message",
    inputSchema: { message: z.string().describe("The message to echo back") }
  }, async ({ message }) => ({
    content: [{ type: "text", text: `Echo: ${message}` }]
  }));

  return server;
}
module.exports = { createMcpServer };

Web Search Server (SerpAPI)

A server that provides Google search, news, and Reddit search:

const { McpServer } = require("@modelcontextprotocol/sdk/server/mcp.js");
const { z } = require("zod");

function createMcpServer() {
  const server = new McpServer({ name: "serpapi", version: "1.0.0" });

  server.registerTool("google_search", {
    title: "Google Search",
    description: "Search Google and get structured results",
    inputSchema: {
      query: z.string().describe("Search query"),
      num: z.number().optional().describe("Number of results (default: 10)")
    }
  }, async ({ query, num = 10 }) => {
    const apiKey = process.env.SERPAPI_KEY;
    if (!apiKey) {
      return { content: [{ type: "text", text: "Error: SERPAPI_KEY not configured" }] };
    }

    const url = `https://serpapi.com/search.json?q=${encodeURIComponent(query)}&num=${num}&api_key=${apiKey}`;
    const response = await fetch(url);
    const data = await response.json();

    const results = (data.organic_results || []).map((r, i) => ({
      position: i + 1,
      title: r.title,
      link: r.link,
      snippet: r.snippet
    }));

    return {
      content: [{ type: "text", text: JSON.stringify({ query, resultCount: results.length, results }) }]
    };
  });

  return server;
}
module.exports = { createMcpServer };

Document Analysis Server (AWS Textract)

Extract text, tables, and forms from documents:

const { McpServer } = require("@modelcontextprotocol/sdk/server/mcp.js");
const { z } = require("zod");
const https = require("https");

function createMcpServer() {
  const server = new McpServer({ name: "aws-textract", version: "1.0.0" });

  server.registerTool("detect_text", {
    title: "Detect Text",
    description: "Extract text from a document image using AWS Textract",
    inputSchema: {
      document_bytes: z.string().describe("Base64-encoded document (PNG, JPEG, PDF)")
    }
  }, async ({ document_bytes }) => {
    // Uses AWS SDK via the platform's credentials
    const AWS = require("aws-sdk");
    const textract = new AWS.Textract({ region: "us-east-1" });

    const result = await textract.detectDocumentText({
      Document: { Bytes: Buffer.from(document_bytes, "base64") }
    }).promise();

    const lines = result.Blocks
      .filter(b => b.BlockType === "LINE")
      .map(b => ({ text: b.Text, confidence: b.Confidence }));

    return {
      content: [{ type: "text", text: JSON.stringify({
        text: lines.map(l => l.text).join("\n"),
        lineCount: lines.length,
        lines
      })}]
    };
  });

  return server;
}
module.exports = { createMcpServer };

Server with Resources and Prompts

A complete server showing all three MCP capabilities:

const { McpServer } = require("@modelcontextprotocol/sdk/server/mcp.js");
const { z } = require("zod");

function createMcpServer() {
  const server = new McpServer({ name: "complete-example", version: "1.0.0" });

  // Tool
  server.registerTool("calculate", {
    title: "Calculate",
    description: "Evaluate a math expression",
    inputSchema: { expression: z.string().describe("Math expression like '2 + 2'") }
  }, async ({ expression }) => {
    try {
      const result = Function('"use strict"; return (' + expression + ')')();
      return { content: [{ type: "text", text: JSON.stringify({ expression, result }) }] };
    } catch (e) {
      return { content: [{ type: "text", text: `Error: ${e.message}` }] };
    }
  });

  // Resource
  server.resource(
    "capabilities",
    "complete-example://capabilities",
    { description: "List of server capabilities", mimeType: "application/json" },
    async (uri) => ({
      contents: [{
        uri: uri.href,
        mimeType: "application/json",
        text: JSON.stringify({
          tools: ["calculate"],
          prompts: ["math-tutor"],
          version: "1.0.0"
        })
      }]
    })
  );

  // Prompt
  server.prompt(
    "math-tutor",
    "Get help solving a math problem step by step",
    { problem: z.string().describe("The math problem to solve") },
    async ({ problem }) => ({
      messages: [{
        role: "user",
        content: {
          type: "text",
          text: `Please solve this math problem step by step, showing your work:\n\n${problem}\n\nUse the calculate tool to verify your answer.`
        }
      }]
    })
  );

  return server;
}
module.exports = { createMcpServer };

Tips for Building Servers

1. Start simple — Get one tool working, then add more
2. Test with curl before connecting to Claude/Cline
3. Check for API keys early and return clear error messages
4. Use JSON.stringify() for structured responses
5. Add .describe() to all Zod schema fields — the AI reads these