How to integrate Census bureau MCP with Vercel AI SDK v6

Trusted by
AWS
Glean
Zoom
Airtable

30 min · no commitment · see it on your stack

Census bureau logo
Vercel AI SDK logo
divider

Introduction

This guide walks you through connecting Census bureau to Vercel AI SDK v6 using the Composio tool router. By the end, you'll have a working Census bureau agent that can get latest population estimate for los angeles county, list top industries in texas by employment, fetch 5-year acs median income for chicago through natural language commands.

This guide will help you understand how to give your Vercel AI SDK agent real control over a Census bureau account through Composio's Census bureau MCP server.

Before we dive in, let's take a quick look at the key ideas and tools involved.

Also integrate Census bureau with

TL;DR

Here's what you'll learn:
  • How to set up and configure a Vercel AI SDK agent with Census bureau integration
  • Using Composio's Tool Router to dynamically load and access Census bureau tools
  • Creating an MCP client connection using HTTP transport
  • Building an interactive CLI chat interface with conversation history management
  • Handling tool calls and results within the Vercel AI SDK framework

What is Vercel AI SDK?

The Vercel AI SDK is a TypeScript library for building AI-powered applications. It provides tools for creating agents that can use external services and maintain conversation state.

Key features include:

  • streamText: Core function for streaming responses with real-time tool support
  • MCP Client: Built-in support for Model Context Protocol via @ai-sdk/mcp
  • Step Counting: Control multi-step tool execution with stopWhen: stepCountIs()
  • OpenAI Provider: Native integration with OpenAI models

What is the Census bureau MCP server, and what's possible with it?

The Census bureau MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to Census Bureau data resources. It provides structured and secure access to a wide range of U.S. demographic, business, and community datasets, so your agent can retrieve population statistics, analyze survey results, fetch business patterns, and explore census variables on your behalf.

  • Retrieve up-to-date population estimates: Have your agent pull the latest demographic and population data for specific states, counties, or cities using the Population Estimates Program (PEP).
  • Analyze American Community Survey results: Access detailed ACS 1-year and 5-year estimates for any geography, helping you understand community trends, housing, and economic data.
  • Explore business statistics by region: Automatically fetch County Business Patterns (CBP) and Annual Business Survey (ABS) data to examine local industry and employment trends.
  • Access decennial census data: Let your agent retrieve variables and statistics from the decennial census by vintage and dataset for deep historical and demographic analysis.
  • Investigate variable metadata and definitions: Effortlessly obtain detailed information about any census variable, including descriptions, data types, and valid values for more informed analysis.

Supported Tools & Triggers

Tools
Geocode AddressTool to geocode a single address to get latitude/longitude coordinates.
Geocode Address for Census GeographiesGeocode an address and return Census geography identifiers including state, county, tract, block group, and block FIPS codes.
Geocode Address PartsTool to geocode an address using separate components (street, city, state, ZIP) to get latitude/longitude coordinates.
Geocode Address with GeographyTool to geocode an address and return both coordinates and Census geography information.
Geocode CoordinatesReverse geocode latitude/longitude coordinates to Census geographic areas.
Geocode Puerto Rico Address with GeographyTool to geocode a Puerto Rico address and return coordinates plus Census geography data.
Batch Geocode Addresses with GeographiesBatch geocode multiple addresses from a CSV file and return Census geography codes.
Geocode Puerto Rico AddressTool to geocode a Puerto Rico address with urbanization to latitude/longitude coordinates.
Get ACS 1-Year EstimatesTool to retrieve 1-year American Community Survey (ACS) estimates for a specified geography.
Get ACS 5-Year EstimatesRetrieve 5-year American Community Survey (ACS) estimates from the U.
Get Community Resilience EstimatesRetrieve U.
Get County Business PatternsTool to retrieve County Business Patterns (CBP) data for a specified year.
Get Dataset Examples HTMLTool to retrieve example queries for a Census dataset in HTML format.
Get Dataset Examples JSONTool to retrieve example API query patterns for a specific Census dataset and vintage.
Get Dataset Examples (XML)Tool to retrieve example queries for a Census Bureau dataset in XML format.
Get Dataset Geography HTMLTool to retrieve available geographies for a Census dataset in HTML format.
Get Dataset Geography JSONTool to get the list of supported geography levels for a specific Census dataset with their hierarchy and required predicates.
Get Dataset Geography XMLTool to retrieve available geographies for a Census Bureau dataset in XML format.
Get Dataset GroupsTool to retrieve the list of table groups for a Census dataset and vintage.
Get Dataset SortsTool to list available sort options for a specific Census dataset and vintage.
Get Dataset TagsTool to list available tags/keywords for a specific Census dataset and vintage.
Get Dataset Variables JSONTool to retrieve the complete list of available variables for a specific Census dataset.
Get Decennial Census DataRetrieve Decennial Census data (population, demographics, housing) from the U.
Get Planning Database DataGet Planning Database (PDB) data containing Census tract and block group level data useful for planning.
Get Population EstimatesRetrieves Population Estimates Program (PEP) data from the US Census Bureau API.
Get TIGERweb ACS Generalized BoundariesTool to access generalized ACS (American Community Survey) boundary services from TIGERweb for specific survey years (2012-2024).
Get TIGERweb Map Service MetadataTool to retrieve TIGERweb MapServer service metadata including available layers, capabilities, and spatial reference information.
Get Timeseries Examples HTMLTool to retrieve HTML-formatted example queries for a Census Bureau timeseries dataset.
Get Timeseries Examples JSONTool to get example queries for a timeseries dataset in JSON format.
Get Timeseries Examples XMLTool to retrieve example queries for a Census Bureau timeseries dataset in XML format.
Get Timeseries Geography HTMLTool to retrieve available FIPS geographies for a timeseries dataset in HTML format.
Get Timeseries Geography JSONTool to get available geographies for a timeseries dataset in JSON format.
Get Timeseries Geography XMLTool to retrieve available geographies for a Census Bureau timeseries dataset in XML format.
Get Timeseries Variables HTMLTool to retrieve a list of available variables for a Census timeseries dataset in HTML format.
Get Timeseries Variables JSONTool to get a list of variables available for a timeseries dataset in JSON format.
Get Timeseries Variables XMLTool to get a list of variables available for a timeseries dataset in XML format.
Get Variable DetailsTool to retrieve metadata for a specific variable in a Census dataset for a given year.
List Available DatasetsLists all available Census Bureau datasets with their metadata, vintages, and API endpoints.
List Datasets HTMLTool to retrieve a complete HTML listing of all available (non-timeseries) Census Bureau datasets.
List Datasets XMLTool to retrieve a list of all available Census Bureau datasets in XML format.
List Geocoder BenchmarksList all available benchmark versions for the Census Bureau geocoding service.
List Geocoder VintagesTool to list available geography vintages for a given Census geocoder benchmark.
List TIGERweb ServicesTool to discover all available TIGERweb map services for Census geographic boundaries.
List Timeseries Datasets (HTML)Tool to retrieve a list of all available timeseries datasets from the US Census Bureau API in HTML format.
List Timeseries Datasets (JSON)Tool to list all available timeseries datasets from the US Census Bureau API.
List Timeseries Datasets (XML)Tool to retrieve a list of all available Census Bureau timeseries datasets in XML format.
Query ACS Supplemental EstimatesQuery ACS Supplemental Estimates data by variables and geography.
Query ACS Comparison ProfilesQuery ACS Comparison Profiles data by variables and geography.
Query ACS Migration FlowsTool to query American Community Survey (ACS) Migration Flows data by variables and geography.
Query ACS Data ProfileTool to query ACS Data Profiles by variables and geography.
Query ACS Selected Population ProfilesTool to query ACS Selected Population Profiles (SPP) data by variables and geography for specific population groups.
Query ACS Subject TablesTool to query ACS Subject Tables data by variables and geography.
Query Annual Business SurveyTool to query Annual Business Survey Company Summary (abscs) data with demographic filters.
Query Commodity Flow SurveyQuery Commodity Flow Survey data on freight shipments by origin, destination, mode, and commodity.
Query CPS Survey DataTool to query Current Population Survey (CPS) microdata including basic monthly employment data and supplemental surveys.
Query Decennial DHCTool to query Decennial Census Demographic and Housing Characteristics (DHC) data by variables and geography.
Query Decennial Census Demographic ProfileTool to query Decennial Census Demographic Profile data by variables and geography.
Query Decennial Census P.L. Redistricting DataTool to query Decennial Census P.
Query Economic Census DataTool to query Economic Census data including establishments, employment, payroll, and receipts by geography and industry (NAICS).
Query International Trade TimeseriesTool to query International Trade timeseries data from Census Bureau API.
Query Nonemployer StatisticsTool to query Nonemployer Statistics data covering businesses with no paid employees.
Query PEP CharAgeGroupsQuery population estimates by age groups, sex, race, and Hispanic origin from the Census Bureau PEP CharAgeGroups dataset.
Query PEP ComponentsQuery components of population change from the Census Bureau Population Estimates Program (PEP).
Query PEP Housing EstimatesQuery housing unit estimates from the US Census Bureau Population Estimates Program (PEP).
Query Population ProjectionsQuery population projections from the Census Bureau API.
Query Surname DataQuery surname frequency data from the U.
Query TIGERweb LayerTool to query TIGERweb GeoServices for Census geographic boundaries and features.
Query Business Dynamics StatisticsQuery Business Dynamics Statistics (BDS) time series data from the Census Bureau.
Query Timeseries DataQuery Census timeseries datasets containing longitudinal data for multiple time periods.
Query Economic Indicators Time SeriesTool to query Economic Indicators Time Series (EITS) data from the US Census Bureau.
Query Residential Construction StatsTool to query Residential Construction statistics from Census Bureau Economic Indicators Time Series (EITS).
Query Residential Sales DataQuery Residential Sales statistics from Census Bureau's Economic Indicator Time Series (EITS).
Query Health Insurance EstimatesQuery Small Area Health Insurance Estimates (SAHIE) from the Census Bureau timeseries API.
Query Household Pulse Survey TimeseriesTool to query Household Pulse Survey (HPS) timeseries data measuring household experiences during the COVID-19 pandemic.
Query International DatabaseQuery International Database (IDB) demographic data for 227 countries and areas worldwide.
Query Timeseries International Trade Exports by HSTool to query international trade exports by Harmonized System code from Census Bureau time series API.
Query Timeseries International Trade Imports by End UseQuery international trade imports by end-use category from Census Bureau timeseries data.
Query Timeseries PovertyQuery poverty statistics from the Census Bureau's timeseries poverty datasets.
Query QWI Timeseries DataQuery Quarterly Workforce Indicators (QWI) timeseries data on employment, earnings, and job flows.
Query Timeseries QWI State/AreaQuery Quarterly Workforce Indicators (QWI) State/Area characteristics from the Census Bureau's time series API.
Query ZIP Business PatternsTool to query ZIP Code Business Patterns (ZBP) data including establishments and employment by ZIP code and industry.

What is the Composio tool router, and how does it fit here?

What is Composio SDK?

Composio's Composio SDK helps agents find the right tools for a task at runtime. You can plug in multiple toolkits (like Gmail, HubSpot, and GitHub), and the agent will identify the relevant app and action to complete multi-step workflows. This can reduce token usage and improve the reliability of tool calls. Read more here: Getting started with Composio SDK

The tool router generates a secure MCP URL that your agents can access to perform actions.

How the Composio SDK works

The Composio SDK follows a three-phase workflow:

  1. Discovery: Searches for tools matching your task and returns relevant toolkits with their details.
  2. Authentication: Checks for active connections. If missing, creates an auth config and returns a connection URL via Auth Link.
  3. Execution: Executes the action using the authenticated connection.

Step-by-step Guide

Prerequisites

Before you begin, make sure you have:
  • Node.js and npm installed
  • A Composio account with API key
  • An OpenAI API key

Getting API Keys for OpenAI and Composio

OpenAI API Key
  • Go to the OpenAI dashboard and create an API key. You'll need credits to use the models, or you can connect to another model provider.
  • Keep the API key safe.
Composio API Key
  • Log in to the Composio dashboard.
  • Navigate to your API settings and generate a new API key.
  • Store this key securely as you'll need it for authentication.

Install required dependencies

bash
npm install @ai-sdk/openai @ai-sdk/mcp @composio/core ai dotenv

First, install the necessary packages for your project.

What you're installing:

  • @ai-sdk/openai: Vercel AI SDK's OpenAI provider
  • @ai-sdk/mcp: MCP client for Vercel AI SDK
  • @composio/core: Composio SDK for tool integration
  • ai: Core Vercel AI SDK
  • dotenv: Environment variable management

Set up environment variables

bash
OPENAI_API_KEY=your_openai_api_key_here
COMPOSIO_API_KEY=your_composio_api_key_here
COMPOSIO_USER_ID=your_user_id_here

Create a .env file in your project root.

What's needed:

  • OPENAI_API_KEY: Your OpenAI API key for GPT model access
  • COMPOSIO_API_KEY: Your Composio API key for tool access
  • COMPOSIO_USER_ID: A unique identifier for the user session

Import required modules and validate environment

typescript
import "dotenv/config";
import { openai } from "@ai-sdk/openai";
import { Composio } from "@composio/core";
import * as readline from "readline";
import { streamText, type ModelMessage, stepCountIs } from "ai";
import { createMCPClient } from "@ai-sdk/mcp";

const composioAPIKey = process.env.COMPOSIO_API_KEY;
const composioUserID = process.env.COMPOSIO_USER_ID;

if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set");
if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set");
if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set");

const composio = new Composio({
  apiKey: composioAPIKey,
});
What's happening:
  • We're importing all necessary libraries including Vercel AI SDK's OpenAI provider and Composio
  • The dotenv/config import automatically loads environment variables
  • The MCP client import enables connection to Composio's tool server

Create Tool Router session and initialize MCP client

typescript
async function main() {
  // Create a tool router session for the user
  const session = await composio.create(composioUserID!, {
    toolkits: ["census_bureau"],
  });

  const mcpUrl = session.mcp.url;
What's happening:
  • We're creating a Tool Router session that gives your agent access to Census bureau tools
  • The create method takes the user ID and specifies which toolkits should be available
  • The returned mcp object contains the URL and authentication headers needed to connect to the MCP server
  • This session provides access to all Census bureau-related tools through the MCP protocol

Connect to MCP server and retrieve tools

typescript
const mcpClient = await createMCPClient({
  transport: {
    type: "http",
    url: mcpUrl,
    headers: session.mcp.headers, // Authentication headers for the Composio MCP server
  },
});

const tools = await mcpClient.tools();
What's happening:
  • We're creating an MCP client that connects to our Composio Tool Router session via HTTP
  • The mcp.url provides the endpoint, and mcp.headers contains authentication credentials
  • The type: "http" is important - Composio requires HTTP transport
  • tools() retrieves all available Census bureau tools that the agent can use

Initialize conversation and CLI interface

typescript
let messages: ModelMessage[] = [];

console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
console.log(
  "Ask any questions related to census_bureau, like summarize my last 5 emails, send an email, etc... :)))\n",
);

const rl = readline.createInterface({
  input: process.stdin,
  output: process.stdout,
  prompt: "> ",
});

rl.prompt();
What's happening:
  • We initialize an empty messages array to maintain conversation history
  • A readline interface is created to accept user input from the command line
  • Instructions are displayed to guide the user on how to interact with the agent

Handle user input and stream responses with real-time tool feedback

typescript
rl.on("line", async (userInput: string) => {
  const trimmedInput = userInput.trim();

  if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) {
    console.log("\nGoodbye!");
    rl.close();
    process.exit(0);
  }

  if (!trimmedInput) {
    rl.prompt();
    return;
  }

  messages.push({ role: "user", content: trimmedInput });
  console.log("\nAgent is thinking...\n");

  try {
    const stream = streamText({
      model: openai("gpt-5"),
      messages,
      tools,
      toolChoice: "auto",
      stopWhen: stepCountIs(10),
      onStepFinish: (step) => {
        for (const toolCall of step.toolCalls) {
          console.log(`[Using tool: ${toolCall.toolName}]`);
          }
          if (step.toolCalls.length > 0) {
            console.log(""); // Add space after tool calls
          }
        },
      });

      for await (const chunk of stream.textStream) {
        process.stdout.write(chunk);
      }

      console.log("\n\n---\n");

      // Get final result for message history
      const response = await stream.response;
      if (response?.messages?.length) {
        messages.push(...response.messages);
      }
    } catch (error) {
      console.error("\nAn error occurred while talking to the agent:");
      console.error(error);
      console.log(
        "\nYou can try again or restart the app if it keeps happening.\n",
      );
    } finally {
      rl.prompt();
    }
  });

  rl.on("close", async () => {
    await mcpClient.close();
    console.log("\n👋 Session ended.");
    process.exit(0);
  });
}

main().catch((err) => {
  console.error("Fatal error:", err);
  process.exit(1);
});
What's happening:
  • We use streamText instead of generateText to stream responses in real-time
  • toolChoice: "auto" allows the model to decide when to use Census bureau tools
  • stopWhen: stepCountIs(10) allows up to 10 steps for complex multi-tool operations
  • onStepFinish callback displays which tools are being used in real-time
  • We iterate through the text stream to create a typewriter effect as the agent responds
  • The complete response is added to conversation history to maintain context
  • Errors are caught and displayed with helpful retry suggestions

Complete Code

Here's the complete code to get you started with Census bureau and Vercel AI SDK:

typescript
import "dotenv/config";
import { openai } from "@ai-sdk/openai";
import { Composio } from "@composio/core";
import * as readline from "readline";
import { streamText, type ModelMessage, stepCountIs } from "ai";
import { createMCPClient } from "@ai-sdk/mcp";

const composioAPIKey = process.env.COMPOSIO_API_KEY;
const composioUserID = process.env.COMPOSIO_USER_ID;

if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set");
if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set");
if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set");

const composio = new Composio({
  apiKey: composioAPIKey,
});

async function main() {
  // Create a tool router session for the user
  const session = await composio.create(composioUserID!, {
    toolkits: ["census_bureau"],
  });

  const mcpUrl = session.mcp.url;

  const mcpClient = await createMCPClient({
    transport: {
      type: "http",
      url: mcpUrl,
      headers: session.mcp.headers, // Authentication headers for the Composio MCP server
    },
  });

  const tools = await mcpClient.tools();

  let messages: ModelMessage[] = [];

  console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
  console.log(
    "Ask any questions related to census_bureau, like summarize my last 5 emails, send an email, etc... :)))\n",
  );

  const rl = readline.createInterface({
    input: process.stdin,
    output: process.stdout,
    prompt: "> ",
  });

  rl.prompt();

  rl.on("line", async (userInput: string) => {
    const trimmedInput = userInput.trim();

    if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) {
      console.log("\nGoodbye!");
      rl.close();
      process.exit(0);
    }

    if (!trimmedInput) {
      rl.prompt();
      return;
    }

    messages.push({ role: "user", content: trimmedInput });
    console.log("\nAgent is thinking...\n");

    try {
      const stream = streamText({
        model: openai("gpt-5"),
        messages,
        tools,
        toolChoice: "auto",
        stopWhen: stepCountIs(10),
        onStepFinish: (step) => {
          for (const toolCall of step.toolCalls) {
            console.log(`[Using tool: ${toolCall.toolName}]`);
          }
          if (step.toolCalls.length > 0) {
            console.log(""); // Add space after tool calls
          }
        },
      });

      for await (const chunk of stream.textStream) {
        process.stdout.write(chunk);
      }

      console.log("\n\n---\n");

      // Get final result for message history
      const response = await stream.response;
      if (response?.messages?.length) {
        messages.push(...response.messages);
      }
    } catch (error) {
      console.error("\nAn error occurred while talking to the agent:");
      console.error(error);
      console.log(
        "\nYou can try again or restart the app if it keeps happening.\n",
      );
    } finally {
      rl.prompt();
    }
  });

  rl.on("close", async () => {
    await mcpClient.close();
    console.log("\n👋 Session ended.");
    process.exit(0);
  });
}

main().catch((err) => {
  console.error("Fatal error:", err);
  process.exit(1);
});

Conclusion

You've successfully built a Census bureau agent using the Vercel AI SDK with streaming capabilities! This implementation provides a powerful foundation for building AI applications with natural language interfaces and real-time feedback.

Key features of this implementation:

  • Real-time streaming responses for a better user experience with typewriter effect
  • Live tool execution feedback showing which tools are being used as the agent works
  • Dynamic tool loading through Composio's Tool Router with secure authentication
  • Multi-step tool execution with configurable step limits (up to 10 steps)
  • Comprehensive error handling for robust agent execution
  • Conversation history maintenance for context-aware responses

You can extend this further by adding custom error handling, implementing specific business logic, or integrating additional Composio toolkits to create multi-app workflows.

How to build Census bureau MCP Agent with another framework

FAQ

What are the differences in Tool Router MCP and Census bureau MCP?

With a standalone Census bureau MCP server, the agents and LLMs can only access a fixed set of Census bureau tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Census bureau and many other apps based on the task at hand, all through a single MCP endpoint.

Can I use Tool Router MCP with Vercel AI SDK v6?

Yes, you can. Vercel AI SDK v6 fully supports MCP integration. You get structured tool calling, message history handling, and model orchestration while Tool Router takes care of discovering and serving the right Census bureau tools.

Can I manage the permissions and scopes for Census bureau while using Tool Router?

Yes, absolutely. You can configure which Census bureau scopes and actions are allowed when connecting your account to Composio. You can also bring your own OAuth credentials or API configuration so you keep full control over what the agent can do.

How safe is my data with Composio Tool Router?

All sensitive data such as tokens, keys, and configuration is fully encrypted at rest and in transit. Composio is SOC 2 Type 2 compliant and follows strict security practices so your Census bureau data and credentials are handled as safely as possible.

Used by agents from

Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai

Never worry about agent reliability

We handle tool reliability, observability, and security so you never have to second-guess an agent action.