MCP App Store
by KeenEthics
Intuit TurboTax icon

Intuit TurboTax MCP Server for Claude and ChatGPT

By Intuit

Overview

Get fast federal tax estimates and discover filing and preparation options with Intuit TurboTax-powered tools. Use a simple tax estimator or a guided interview to collect what matters, and watch your refund or amount owed update as you go. Get a clear view of the forms and documents you need, what you have, and what is left to gather—then upload securely when you sign in and complete consent when prompted for protected features. Ready for a human? Find trusted tax pros, check real availability, and book an appointment. Confident tax filing starts here.

How to connect

  1. Claude: Open the Intuit TurboTax listing in the Claude connectors directory and choose Connect. Claude listing Intuit TurboTax exposes 3 tools here.
  2. ChatGPT: Open the Intuit TurboTax listing in the ChatGPT app directory and choose Connect. ChatGPT listing Intuit TurboTax exposes 11 tools here.

Tools

Intuit TurboTax tools

ChatGPT
Get fast federal tax estimates and discover filing and preparation options with Intuit TurboTax-powered tools. Use a simple tax estimator or a guided interview to collect what matters, and watch your refund or amount owed update as you go. Get a clear view of the forms and documents you need, what you have, and what is left to gather—then upload securely when you sign in and complete consent when prompted for protected features. Ready for a human? Find trusted tax pros, check real availability, and book an appointment. Confident tax filing starts here.

book_appointment

ChatGPT
Use this when the user wants to book an appointment with a tax expert they've viewed. REQUIRES a prior expert_search call in this conversation. REQUIRED WORKFLOW: 1. expert_search must have been called first 2. From the search response's structuredContent, extract: - search_id (from structuredContent.search_id) - id (from structuredContent.experts[].id for the chosen expert) - availability_id (from structuredContent.experts[].closest_availability[].slots[].id for the time slot) 3. Collect from user: - first_name: User's first name - last_name: User's last name - phone_number: User's phone number (should be numbers only - you can remove formatting characters like +, -, ., spaces, or parentheses before passing) - email: User's email address 4. MANDATORY CONFIRMATION - You MUST explicitly confirm ALL booking details with the user and wait for their response: Display these details clearly: - Name (first and last) - Phone number - Email address - Expert name - Date and time of appointment Then ask: 'Please confirm these details are correct before I proceed with the booking.' 5. Only after user explicitly confirms (e.g., 'yes', 'looks good', 'confirm'), call this tool CRITICAL RULES: - NEVER skip the confirmation step - you MUST always confirm details and wait for user response before booking - ONLY use search_id, id, and availability_id from expert_search's structuredContent - NEVER make up or guess these values - NEVER show IDs (search_id, availability_id, etc.) to the user in conversation - If you don't have a recent expert_search result, call expert_search first - The availability_id must be from the specific time slot the user wants to book RETRY BEHAVIOR: - If the booking did not go through for any reason and the user wants to retry, you MAY retry the booking with the same parameters or different parameters DO NOT use this tool for: - Searching for experts (use expert_search instead) - Modifying or canceling bookings

document_upload_private

ChatGPT
Internal tool for uploading files. Called by the file upload widget with all files in a single call. Accepts a JSON array of file objects. For small static content files like tax forms.

expert_availability

ChatGPT
Fetch detailed availability for a specific expert. Use this tool when a user wants to see more time slots for an expert or schedule an appointment at a specific time. PARAMETERS: - search_id: REQUIRED. The search_id from the previous expert_search call - date: Optional date filter (YYYY-MM-DD). If not provided, returns all available slots RETURNS: - Expert basic info (name, photo, rating, experience) - Available time slots grouped by date - Widget UI for slot selection IMPORTANT: - DO NOT list specific times or time slots in your response to the user - DO NOT show internal IDs (search_id, availability_id, etc.) to the user - Let the widget display all time options interactively - Simply tell the user to select a time from the widget

interview

ChatGPT
Conduct a structured multi-page tax interview. Each user has one active session — no session ID needed. INVALID COMBINATIONS — these are bugs: - topic + anything else (answers, action, bulk_entry, etc.): topic only starts a session; no other params are valid in the same call - answers with invented field names: field_id values must be copied from the tool response, never guessed - any key other than field_id in an answers entry (not "key", "name", "id", "question", "answer", "fieldId", etc.) - any parameter not in this list: topic, answers, action, bulk_entry REMOVING OR CHANGING A SITUATION: There is no delete or remove operation. To undo a situation (e.g., user says "actually I don't have student loans"): interview(answers=[{"field_id": "<copied from response>", "value": false}], bulk_entry=True) Set the field to false — do not try to omit, delete, or use a remove_items key. MODE SELECTION — choose exactly one: | Goal | Params | |------------------------------------|---------------------------------| | Start new interview | topic only | | Start new interview + view all fields | topic only, then call again with bulk_entry=True | | Peek at current page | (no params) | | Save answers, stay on page | answers only | | Save answers and advance | answers + action | | Go back (no validation) | action="back" | | Complete interview | action="complete" | | Peek all fields (all pages) | bulk_entry=True | | Bulk save many fields | answers + bulk_entry=True | To save progress and exit mid-interview: submit answers without action — answers are persisted immediately. Resume any time by calling interview() with no parameters. PARAMETERS: - topic (string): Interview topic. Starts a new session, replacing any existing one. NEVER combine with answers, action, or bulk_entry. - answers (array): Each entry must be {"field_id": "...", "value": ...}. The key is always field_id. Copy field_id exactly from the returned fields list — never invent field names. Field types (boolean, string, number) are indicated in each field's definition in the response. For yes/no fields use JSON booleans (true/false), not strings. - action (string): Copy the action_id value from the returned actions list. Valid values are page-specific — passing an invalid action returns an error listing valid options. Answers submitted alongside action are persisted before the action executes. - bulk_entry (boolean, default false): NOT persisted — set true on every call where bulk behavior is needed. AUTO-MERGE: Applies ONLY when action != "back" AND bulk_entry is false. When active: required fields that already have a current_value are preserved even if omitted from answers. When NOT active (action="back" or bulk_entry=True): only the fields you explicitly include in answers are updated. Fields you omit are not touched. UPDATING A SPECIFIC SITUATION (e.g. "I have mortgage interest", "I got married"): You do not know the field_ids in advance — you must fetch them first. Always use this two-step flow: Step 1 — start interview and get all fields (two separate calls): interview(topic="tax_situation") interview(bulk_entry=True) → returns ALL fields; read each field's field_id from the response Step 2 — submit using field_ids copied from the step 1 response: interview(answers=[{"field_id": "<copied from step 1 response>", "value": true}, ...], bulk_entry=True) EXAMPLES: PAGE-BY-PAGE FLOW (typical linear interview): Step 1 — start, get first page interview(topic="tax_situation") Step 2 — collect answers from user, submit and advance interview(answers=[{"field_id": "filing_status", "value": "single"}, {"field_id": "had_w2", "value": true}], action="next") Step 3 — repeat for each page until is_complete=True interview(answers=[{"field_id": "wages", "value": 85000}], action="next") Peek at current page (resume after a break) interview() Go back interview(action="back") Bulk save across pages (no auto-merge, no navigation) interview(answers=[{"field_id": "had_w2", "value": true}, {"field_id": "wages", "value": 85000}], bulk_entry=True) BEHAVIORAL RULES: - Never invent field…

search_tax_knowledge

ChatGPT
Searches authoritative tax-knowledge sources (IRS and IGC) for content relevant to a tax question, using RAG over curated knowledge bases. USE THIS WHEN - A Skill needs grounded source passages to answer a tax question. - You want citations or supporting text for tax guidance. DO NOT INVOKE - For refund status (use the refund tracker tool). - For tax estimation or calculation (use the tax estimate tool).

tax_checklist

ChatGPT
Display the tax document checklist showing required and completed documents. USE THIS WHEN: - User wants to see what tax documents are needed - User asks about their tax document checklist or status - User wants to know what documents they still need to provide UPDATING TAX SITUATION DATA: If the user's current message contains new tax situation information (e.g., mentions a new job, a new dependent, rental income, mortgage interest, or any other change), update that data using the interview tool before calling this tool. To update a single field: call interview(topic='tax_situation'), then interview(bulk_entry=True) to get field_ids, then interview(answers=[{"field_id": "<id>", "value": ...}], bulk_entry=True) to submit. To update multiple fields or complete a full interview: call interview(topic='tax_situation') and step through all pages. Call this tool after either flow completes. Users can always change their answers (e.g., 'I don't have a W-2 after all'). Use the interview tool to set the relevant field to false — the checklist will update to reflect the new situation. UPLOADED DOCUMENTS -- IMPORTANT: - Once a document has been uploaded, it CANNOT be deleted, removed, or modified through this tool or through chat. - There are no 'delete', 'remove', or 'undo upload' actions available. - If the user asks to delete or remove an uploaded document, tell them: 'Uploaded documents cannot be removed here. To manage uploaded files, use Continue in TurboTax.' - Never suggest or imply that uploaded documents can be deleted via chat. PARAMETER RULES: - DO NOT set the 'mode' parameter. Always omit it and let it default. - The 'mode' parameter is reserved for internal widget use only. - You must NEVER pass mode='full' unless the user's message explicitly contains the exact phrase 'mode full' or the widget instructs you to. - If you are unsure, do not set mode. The default behavior is correct. This tool supports mixed authentication. If the user has not authenticated, a preview screen is shown with an overview of checklist capabilities and a prompt to connect their TurboTax account. After authentication, the full personalized checklist is returned grouped by category with completion status. AFTER THE WIDGET (AUTHENTICATED FULL CHECKLIST): The tool result includes model-only instructions and neutral facts for a brief follow-up you compose for the user. Do not read the instruction block verbatim. After you show the checklist, speak to the user in a conversational tone (no bullet walls). Cover: (1) categories and items on their checklist, (2) documents already uploaded or connected, (3) an invitation to update in natural language, (4) next steps: they can upload documents in this checklist (per-item or dropzone) or use Continue in TurboTax to continue or file; do not imply uploads only happen in TurboTax (do not paste URLs). Use structuredContent.followUp and the 'facts' / 'prompt' lines in the text as inputs; translate them into natural second-person speech. CONSENT: This tool requires IRC §7216 tax disclosure consent. Do not proceed with this tool's action until consent is confirmed. The user must accept or deny using the buttons in the consent widget -- never via chat. If the user previously declined, that is NOT permanent; the server will show a fresh consent form.

tax_estimate

ChatGPT
IMPORTANT: Only tax year 2025 estimates are supported. If the user asks for a different tax year, inform them that only 2025 estimates are currently available. USE THIS WHEN - the user explicitly requests a US federal tax estimate, refund or amount owed and required minimum info is provided. - If unclear or incomplete, ask clarifying questions before invoking. MINIMUM INFO - Filing status, confirm MFJ vs MFS - Age - Income and withholdings details - If filing status is MFJ or MFS, ask for spouse income and withholdings details. - Whether the tax payer can be claimed as a dependent on another person's tax return. CRITICAL BEHAVIORS (MUST FOLLOW): - DO NOT USE WORLD KNOWLEDGE OR MAKE ASSUMPTIONS. USE ONLY THE CONTENT PROVIDED IN THE TOOL DESCRIPTION. - Do NOT ask the user for the standard deduction amount. This is automatically calculated based on filing status. If its provided, ignore it. DO NOT INVOKE - For general tax questions, non-calculation requests, or when the user declines an estimate. PRIOR ESTIMATES - If a prior estimate exists and inputs haven't changed, reference it. Invoke only if values change or the user requests recalculation. INVOCATION RULES - Do NOT ask the user for the standard deduction amount. This is automatically calculated based on filing status. If its provided, ignore it. - Use only user-provided fields. Omit keys entirely if not provided. Do not send null, empty strings, or 0 unless the user stated 0 explicitly. - Do not invent values. Only pass fields the user provided. Don't pick values from examples. - Accept $ and commas; strip formatting before sending. Preserve cents when provided. - Do not annualize per-paycheck or monthly amounts unless the user provides frequency and you confirm conversion. - Use the latest values; ask if user is providing conflicting information. AFTER INVOKING - After invoking, use the tool response as a driver for follow-up questions. - The tool response includes follow-up questions. ALWAYS ask all the follow-up questions. They are NOT OPTIONAL. - ALWAYS suggest the next_actions included in the tool response. - Do not modify the tool's call-to-action text or link and any other returned values. - Present the summary and offer to re-run if inputs change. - The needs follow up status is included in the tool response. If it is "needs_follow_up", then ask all the follow-up questions and suggest the next actions also.

tax_filing_options

ChatGPT
Use this when user expresses intent to start filing, file taxes, or wants to know filing options. Display TurboTax filing options to help users choose how to file their taxes. Shows DIY, Expert Review, and Full Service options in a carousel format with features and call-to-action buttons. Post-Invocation Guidelines After the carousel is rendered, you must strictly present the next_actions provided in the tool's JSON response that lead to a tax expert consultation for personalized advice

tax_estimate

Claude

tax_filing_options

Claude

Capabilities

WritesInteractive

Example Prompts

Click any prompt to copy it.

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/intuit-turbotax"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/intuit-turbotax)