Connect Google Forms to ChatGPT: Access Forms and Analyze Responses

If you need to connect Google Forms to ChatGPT to automate survey analysis, extract structured data from user submissions, or dynamically query form structures, you need a Model Context Protocol (MCP) server. This server acts as the translation layer between ChatGPT's JSON-RPC tool calls and the underlying Google Forms REST API.

If your team uses Claude, check out our guide on connecting Google Forms to Claude or explore our broader architectural overview on connecting Google Forms to AI Agents.

Giving a Large Language Model (LLM) read access to Google Forms is a massive engineering challenge. You have to handle complex, polymorphic JSON schemas, map disjointed question IDs to answer payloads, and navigate dual OAuth scopes across Google Drive and Google Forms. Every time an LLM attempts to parse a complex survey grid, your custom server code must interpret the API's abstract data models and feed them back to the model without blowing up its context window.

This guide breaks down exactly how to use Truto to generate a secure, managed MCP server for Google Forms, connect it natively to ChatGPT, and execute complex data analysis workflows using natural language.

Stop writing boilerplate API integration code. Let Truto generate secure, managed MCP servers for your AI agents in seconds. :::

The Engineering Reality of the Google Forms API

A custom MCP server is a self-hosted integration layer. While the open MCP standard provides a predictable way for models to discover tools, implementing it against the Google Forms API is exceptionally painful. If you decide to build a custom MCP server, you own the entire lifecycle.

Here are the specific integration challenges that break standard CRUD assumptions when working with Google Forms:

Polymorphic Item Schemas

The Google Forms API does not return a flat list of questions. A Form object contains an items array. Every element in this array is polymorphic. An item could be a questionItem, a questionGroupItem, a pageBreakItem, or an imageItem. Inside a questionItem, the actual question payload branches again into choiceQuestion, textQuestion, scaleQuestion, or gridQuestion. If you try to feed this raw OpenAPI spec directly into an LLM, the model will frequently hallucinate tool arguments because the schema depth exceeds standard reasoning limits.

Disjointed Response Mapping

When an LLM asks "What did users say about our pricing?", a standard API call to fetch responses does not return text-based question-and-answer pairs. Instead, a FormResponse object contains an answers map where the key is an abstract questionId (e.g., "1a2b3c") and the value is the user's input.

To make sense of a submission, your system must first fetch the form definition, traverse the polymorphic tree to find the text string for "1a2b3c", and then map that string to the answer payload. Your MCP server must either handle this relational mapping in memory, or explicitly provide the LLM with separate tools to fetch the form and the responses so the model can correlate the IDs in its own context.

Dual Drive and Forms OAuth Scopes

Google Forms are technically Google Drive files with the MIME type application/vnd.google-apps.form. If you want an LLM to "list all available forms," you cannot simply call the Forms API. You must query the Google Drive API. This means your OAuth application must request and maintain scopes for both Drive (https://www.googleapis.com/auth/drive.readonly) and Forms (https://www.googleapis.com/auth/forms.responses.readonly). Handling token refresh cycles across multiple Google APIs requires durable state management.

How to Generate a Google Forms MCP Server

Instead of building a schema parser and OAuth state machine from scratch, you can use Truto to dynamically generate an MCP server.

Truto derives MCP tools directly from the integration's underlying resource definitions and documentation records. A tool only appears in the MCP server if it has a corresponding documentation entry, acting as a curation mechanism to ensure ChatGPT only sees highly optimized, AI-ready endpoints.

Every MCP server is scoped to a single integrated account (a connected Google instance for a specific tenant) and authenticated via a cryptographic token in the URL.

You can generate this server via the Truto UI or programmatically via the API.

Method 1: Via the Truto UI

1. Navigate to the Integrated Accounts page in your Truto dashboard and select your connected Google Forms account.
2. Click the MCP Servers tab.
3. Click Create MCP Server.
4. Select your desired configuration (e.g., name the server, select read methods, apply any relevant tags, and set an optional expiration date).
5. Click generate and copy the provided MCP server URL (e.g., https://api.truto.one/mcp/a1b2c3d4...).

Method 2: Via the Truto API

For teams building programmatic AI agents, you can generate the MCP server via a single API call. Truto validates the integration's AI-readiness, generates the cryptographic token, stores it in distributed KV storage for low-latency routing, and returns the URL.

Request:

curl -X POST https://api.truto.one/integrated-account/{integrated_account_id}/mcp \
  -H "Authorization: Bearer $TRUTO_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Google Forms Analysis Agent",
    "config": {
      "methods": ["read"],
      "tags": ["forms", "responses"]
    },
    "expires_at": "2026-12-31T23:59:59Z"
  }'

Response:

{
  "id": "mcp_8f7e6d5c",
  "name": "Google Forms Analysis Agent",
  "config": {
    "methods": ["read"],
    "tags": ["forms", "responses"]
  },
  "expires_at": "2026-12-31T23:59:59Z",
  "url": "https://api.truto.one/mcp/a1b2c3d4e5f67890"
}

Treat the url as a sensitive credential. It encodes the tenant routing and authentication required to execute tools against that specific Google account.

Connecting the MCP Server to ChatGPT

Once you have your Truto MCP URL, you can expose the Google Forms tools to ChatGPT. The connection process relies entirely on JSON-RPC 2.0 messages sent over HTTP POST.

Method A: Via the ChatGPT UI

If you are using ChatGPT Pro, Plus, Business, Enterprise, or Education, you can connect the server natively through the web interface:

1. Open ChatGPT and navigate to Settings -> Apps -> Advanced settings.
2. Toggle Developer mode on.
3. Under MCP servers / Custom connectors, click to add a new server.
4. Enter a name (e.g., "Google Forms (Truto)").
5. Paste your Truto MCP URL into the Server URL field.
6. Click Add.

ChatGPT will immediately perform an MCP handshake (initialize), request the available capabilities, and populate the model's context with the Google Forms tools.

Method B: Via Manual Config File (SSE Transport)

If you are wrapping ChatGPT in a custom enterprise agent framework or testing locally with an MCP inspector, you configure the connection using a JSON config file. Because Truto provides a remote HTTP endpoint, you use the standard @modelcontextprotocol/server-sse wrapper to bridge standard stdio transport to the remote Truto URL.

{
  "mcpServers": {
    "google-forms-truto": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-sse",
        "https://api.truto.one/mcp/a1b2c3d4e5f67890"
      ]
    }
  }
}

Hero Tools for Google Forms

Truto translates the complex Google Forms API into flattened, LLM-friendly schemas. Here are the highest-leverage tools available for ChatGPT.

get_single_form_by_id

Retrieves the complete structural definition of a Google Form. This tool is critical because it returns the schema containing all questionId values, title metadata, and layout information. ChatGPT uses this to understand the context of what a form is actually asking before it analyzes responses.

"Fetch the form definition for ID '1xY2z3A4' and list out all the question text along with their corresponding question IDs."

list_all_forms

Lists all Google Forms accessible to the authenticated user. Because Forms are stored in Google Drive, Truto handles the underlying Drive API metadata querying automatically. It returns an array of objects representing forms, including their id, name, and mimeType.

"Search my Google account and list all the Google Forms I have access to, providing their names and file IDs."

list_all_forms_responses

Retrieves all user submissions for a specific Google Form. The LLM must pass the form_id. Truto automatically injects limit and next_cursor properties into the query schema to handle pagination. The LLM will receive the answers map containing abstract IDs, which it can cross-reference with the form definition.

"Get all the responses for the Customer Feedback form (ID '1xY2z3A4'). Paginate through the results if there are more than 50 submissions."

get_single_forms_response_by_id

Fetches a specific, individual submission by its ID. This is highly useful for agentic workflows where a webhook notifies the system of a new submission, and ChatGPT needs to fetch the exact payload of that specific entry to trigger a downstream action.

"Retrieve the specific form response ID 'Resp998877' from the Onboarding form, and summarize the user's answers."

list_all_oauth_user_info

Retrieves basic profile information about the authenticated Google user. This includes their unique identifier, full name, profile picture URL, and email address. Agents use this to verify identity context before taking consequential actions.

"Who is the authenticated Google user right now? Please provide their full name and email address."

For the complete tool inventory and granular JSON schema definitions, visit the Google Forms integration page.

Workflows in Action

When you give ChatGPT access to these curated tools, it can perform multi-step reasoning to bypass the architectural quirks of the Google Forms API.

Scenario 1: Cross-Referencing Form Structure with Survey Results

A product manager wants ChatGPT to analyze sentiment on a recent feature launch survey.

"Find the 'Q3 Feature Launch Survey' form, get all of its responses, and give me a summary of negative feedback regarding the pricing question."

1. ChatGPT calls list_all_forms to locate the exact id of the "Q3 Feature Launch Survey".
2. ChatGPT calls get_single_form_by_id using that ID. It reads the polymorphic items array to locate the specific questionId that corresponds to the text "What are your thoughts on the new pricing?".
3. ChatGPT calls list_all_forms_responses using the form ID.
4. ChatGPT processes the responses, specifically filtering the answers map for the questionId it identified in Step 2.
5. ChatGPT synthesizes the findings and outputs a natural language summary of the negative pricing feedback.

sequenceDiagram
    participant ChatGPT as ChatGPT
    participant TrutoMCP as Truto MCP Server
    participant Upstream as Google APIs
    
    ChatGPT->>TrutoMCP: Call list_all_forms
    TrutoMCP->>Upstream: GET /v3/files (Drive API)<br>q=mimeType='...form'
    Upstream-->>TrutoMCP: Returns file list
    TrutoMCP-->>ChatGPT: Returns form IDs
    
    ChatGPT->>TrutoMCP: Call get_single_form_by_id
    TrutoMCP->>Upstream: GET /v1/forms/{id}
    Upstream-->>TrutoMCP: Returns polymorphic schema
    TrutoMCP-->>ChatGPT: Returns unified schema
    
    ChatGPT->>TrutoMCP: Call list_all_forms_responses
    TrutoMCP->>Upstream: GET /v1/forms/{id}/responses
    Upstream-->>TrutoMCP: Returns answers map
    TrutoMCP-->>ChatGPT: Returns response JSON

Scenario 2: Auditing Specific User Submissions

An HR administrator needs to look up a specific employee's IT request based on a submission ID provided in a Slack message.

"Look up form response 'R_12345' in the IT Hardware Request form (ID 'F_9988'). Tell me what laptop model they requested and who the user is."

1. ChatGPT calls get_single_form_by_id with F_9988 to map the question text ("Select your laptop model") to its underlying questionId.
2. ChatGPT calls get_single_forms_response_by_id with F_9988 and R_12345.
3. ChatGPT matches the IDs, identifies the laptop model from the response payload, and extracts the respondent's email from the response metadata.
4. ChatGPT returns a concise answer to the user.

Security and Access Control

Exposing an enterprise Google environment to an LLM requires strict boundaries. Truto provides several mechanisms on the MCP token to restrict what ChatGPT can do.

• Method Filtering (config.methods): Restrict the MCP server to specific operation types. Setting methods: ["read"] ensures the LLM can only execute get and list operations. Write operations are blocked at the server level.
• Tag Filtering (config.tags): Scope the server to specific functional areas. By passing tags like ["responses"], you ensure ChatGPT only sees tools related to form submissions, hiding unrelated endpoints.
• Expiration (expires_at): Set a strict Time-to-Live (TTL) for the MCP server. Truto enforces this via Cloudflare KV expiration and a Durable Object cleanup alarm, making it ideal for temporary contractor access or ephemeral agent sessions.
• Extra Authentication (require_api_token_auth): By default, possession of the MCP URL grants access. Enabling this flag forces the client to also pass a valid Truto API token in the Authorization header, adding a secondary security layer.

Understanding Rate Limits

When connecting ChatGPT to Google Forms, you are subject to Google's API quotas (e.g., requests per minute for Forms and Drive).

It is a crucial architectural fact that Truto does not retry, throttle, or apply backoff on rate limit errors. When Google returns an HTTP 429 Too Many Requests error, Truto passes that error directly to the caller.

Truto normalizes the upstream rate limit information into standardized headers (ratelimit-limit, ratelimit-remaining, ratelimit-reset) per the IETF specification. Your agent framework, not Truto, is responsible for reading these headers and executing the appropriate retry and backoff logic to prevent overwhelming the Google API.

Wrapping Up

Building a custom integration between ChatGPT and Google Forms forces you to deal with polymorphic schemas, dual-scope OAuth requirements, and abstract relational data mapping.

By leveraging Truto's dynamic MCP server generation, you abstract away the API mechanics. Your LLM receives clean, documented JSON-RPC tools derived directly from live schemas, allowing it to autonomously discover forms, cross-reference IDs, and extract deep insights from user submissions in minutes.