ChatGPT integration

Integration status

This integration is live. The Tallyfy MCP server at https://mcp.tallyfy.com supports OAuth 2.1 with PKCE and works with ChatGPT’s Apps/Connectors feature. Use Streamable HTTP transport - SSE is no longer supported.

Connecting ChatGPT to Tallyfy

Connect your ChatGPT subscription to Tallyfy through the MCP server at https://mcp.tallyfy.com. Once set up, you can search, read, and manage tasks, processes, and templates using plain language - no API keys or code required.

The connection uses OAuth 2.1 with PKCE. ChatGPT handles the auth flow automatically - you just sign in with your Tallyfy credentials and authorize access.

1. Enable Developer Mode in ChatGPT

   Click your profile icon in the top-left corner, then go to Settings > Apps and Connectors > Advanced Settings. Turn on Developer Mode (required for write access to MCP tools - read-only Deep Research works without it).

2. Open the Connectors or Apps section

   In ChatGPT Settings, go to Connectors (or Apps - OpenAI renamed this in December 2025, so the label depends on your version). Click Create or Add Custom App.

3. Configure the connector

   Fill in these details:

   • Name: Tallyfy
   • Description: Connect to Tallyfy to search and manage workflow tasks, processes, and templates
   • MCP Server URL: https://mcp.tallyfy.com
   • Authentication: OAuth 2.1 with PKCE (ChatGPT auto-discovers the endpoints)

4. Authorize with your Tallyfy account

   Click Connect. ChatGPT queries https://mcp.tallyfy.com/.well-known/oauth-protected-resource to discover the OAuth endpoints, then redirects you to Tallyfy’s login page. Sign in and click Authorize to grant access. OpenAI stores your refresh tokens securely.

5. Test the connection

   Start a new conversation and try:

   Show me my open tasks in Tallyfy

   If you see a tool call step in the response followed by your task list, the connection is working.

6. Share with your team (optional)

   Publish the connector to your ChatGPT workspace for team-wide access. Admins control visibility and permissions in workspace settings.

OAuth endpoints reference

ChatGPT auto-discovers these through /.well-known endpoints - no manual configuration needed. For debugging or custom client setup:

Endpoint	URL
MCP server	https://mcp.tallyfy.com
Protected resource metadata (RFC 9728)	https://mcp.tallyfy.com/.well-known/oauth-protected-resource
Authorization server metadata (RFC 8414)	https://mcp.tallyfy.com/.well-known/oauth-authorization-server
Authorization (proxy)	https://mcp.tallyfy.com/mcp/oauth/authorize
Token exchange (proxy)	https://mcp.tallyfy.com/mcp/oauth/token
Client registration (proxy)	https://mcp.tallyfy.com/mcp/oauth/register
Tallyfy auth server	https://account.tallyfy.com

Supported OAuth scopes (ChatGPT requests all of these during authorization):

Scope	Access
mcp.tasks.read	Search and retrieve task data
mcp.tasks.write	Create and update tasks
mcp.processes.read	View workflow process status
mcp.processes.write	Manage process lifecycle
mcp.templates.read	Read template configurations
mcp.templates.write	Edit templates, steps, and automations
mcp.forms.read	View form field definitions
mcp.forms.write	Add and update form fields
mcp.users.read	View users and guests
mcp.users.write	Invite and manage users
mcp.automation.read	View automation rules
mcp.automation.write	Create and modify automations

Available MCP tools

Once connected, ChatGPT can call these Tallyfy tools through natural language:

Task tools

• get_my_tasks - Get all tasks assigned to the current user (no parameters needed)
• get_user_tasks - Get tasks assigned to a specific team member by user ID
• get_tasks_for_process - Get all tasks in a specific process by ID or name
• create_task_from_text - Create a task from natural language - extracts title, deadline, and assigns to named members or guests

Search tools

• search_for_tasks - Search tasks across the organization by keyword
• search_for_processes - Search workflow processes/runs by keyword
• search_for_templates - Search templates by keyword (use get_all_templates for a full list)

Process tools

• get_organization_runs - Get all workflow processes with optional filters for status, template, owner, tags, and type

Template tools

• get_template - Get a template by ID or name with full details
• get_all_templates - Get all templates in the organization
• get_kickoff_fields - Get kick-off form fields for a template
• get_step_dependencies - Analyze which automations control when a step appears
• assess_template_health - Full health check covering step clarity, form fields, automations, and deadlines
• suggest_step_deadline - Get deadline recommendations for a step
• suggest_kickoff_fields - Get suggested kick-off fields based on template content
• add_step_to_template - Add a new step to an existing template
• add_assignees_to_step - Add member or guest assignees to a step
• edit_description_on_step - Update the description/instructions on a step

Form field tools

• add_form_field_to_step - Add text, dropdown, date, or other field types to a step
• update_form_field - Update an existing form field’s properties
• move_form_field - Reorder a form field within a step
• delete_form_field - Remove a form field from a step
• get_dropdown_options - Get the options for a dropdown field
• update_dropdown_options - Update dropdown choices
• suggest_form_fields_for_step - Get AI recommendations for useful form fields based on step content

Automation tools

• create_automation_rule - Create if-then automation rules on a template
• update_automation_rule - Modify an existing automation rule’s conditions or actions
• delete_automation_rule - Remove an automation rule
• analyze_template_automations - Analyze all automations for conflicts and redundancies
• get_step_visibility_conditions - See which automations control a step’s visibility
• suggest_automation_consolidation - Get AI recommendations for simplifying automation rules
• consolidate_automation_rules - Preview or apply automation consolidation changes

User and guest tools

• get_organization_users - Get all organization members with profile data
• get_organization_users_list - Get members with minimal data for listing
• get_organization_guests - Get guest users with full data
• get_organization_guests_list - Get guests with minimal data
• invite_user_to_organization - Invite a new member by email

Write operation security

ChatGPT shows confirmation modals before any write operation (create, update, delete). You always approve changes before they happen.

Real-world usage examples

Finding urgent tasks

You: "Show me all urgent tasks assigned to the sales team"
ChatGPT: Calls search_for_tasks, returns filtered task list with details

Checking process status

You: "What's the status of the employee onboarding process for John Smith?"
ChatGPT: Calls search_for_processes, returns process details, completion %, current step

Template health audit

You: "Analyze our Customer Onboarding template and suggest improvements"
ChatGPT: Chains get_template > assess_template_health > analyze_template_automations >
         suggest_automation_consolidation, then summarizes findings

Bulk task review

You: "Get all my tasks and summarize which ones are overdue"
ChatGPT: Calls get_my_tasks, analyzes due dates, provides organized summary

Creating automation rules

You: "Add a rule to the Sales Template that assigns the contract review step
     to Sarah when deal value exceeds $50,000"
ChatGPT: Calls get_template to find the step ID, then create_automation_rule,
         confirms creation after your approval

Cross-referencing data

You: "Find the onboarding template, get all its form fields, then show me
     active onboarding processes and highlight which fields are most
     commonly left empty"

ChatGPT chains:
1. get_all_templates to find the onboarding template
2. get_template to retrieve form field details
3. search_for_processes to find active runs
4. Analyzes and summarizes completion patterns

Tips for better queries

Be specific with search terms

Good: "Find all tasks with 'invoice approval' in the title assigned to me"
Vague: "Show tasks"

Provide context for filtering

Good: "Get processes from the Customer Onboarding template created in the last 7 days"
Vague: "Find recent processes"

Specify what you want to modify

Good: "Update the 'Priority' dropdown in the Bug Report template to add 'Critical' as an option"
Vague: "Change the priority field"

Ask for structured summaries

Good: "List all my tasks grouped by due date, showing only title and assignee"
Vague: "What are my tasks?"

Limitations of the text interface

ChatGPT’s text-based UI can’t handle everything in Tallyfy:

Visual workflow representation

ChatGPT can’t display Tallyfy’s visual process tracker or workflow diagrams. You won’t see step flows, progress bars, or process dependencies at a glance. For visual monitoring, use Tallyfy’s tracker view directly.

Complex form interactions

Dropdown fields with many options, date pickers, and file uploads don’t translate well to plain text. File attachments can’t be handled at all.

Real-time updates

ChatGPT doesn’t push live updates. Changes made by teammates in Tallyfy won’t appear until you ask again - no live notifications.

Bulk visual operations

Drag-and-drop step reordering, batch update previews, and visual branching structures stay in Tallyfy’s native interface.

Security and data handling

• OAuth 2.1 with PKCE - token-based access with no API keys passed around
• Granular scopes - 12 separate read/write scopes let you control exactly what ChatGPT can access
• Write confirmations - ChatGPT shows modals before any create, update, or delete operation
• Token storage - refresh tokens stored securely by OpenAI, revocable anytime from ChatGPT Settings > Connectors
• No training on your data - OpenAI doesn’t use API interactions for model training (per their usage policies)
• Prompt injection risk - be cautious if your task or template content contains prompt-like instructions and verify unexpected ChatGPT behaviors before acting on them

Troubleshooting

“ChatGPT can’t find the Tallyfy connector”

• Verify Developer Mode is enabled: ChatGPT Settings > Apps and Connectors > Advanced Settings
• Confirm your ChatGPT plan supports Apps/Connectors (Plus, Pro, Team, Enterprise, Education)
• Re-create the connector with the correct URL: https://mcp.tallyfy.com

“Authentication fails or keeps looping”

• Re-authorize from ChatGPT Settings > Connectors > Tallyfy > Reconnect
• Confirm your Tallyfy account is active with at least a Standard user role
• Open https://mcp.tallyfy.com/.well-known/oauth-protected-resource in a browser - a JSON response confirms the server is reachable
• Clear browser cache and retry

“ChatGPT says it can’t access certain tools”

• Verify you approved all OAuth scopes during initial connection
• Some write operations require an Administrator role in Tallyfy
• Try disconnecting and reconnecting to re-request all scopes

“Search results are empty or incomplete”

• Make queries more specific - search_for_tasks requires a non-empty keyword
• Verify the items exist in your Tallyfy workspace and that your user has permission to view them

“SSE connection errors”

• SSE transport is no longer supported - the Tallyfy MCP server uses Streamable HTTP only
• Update to the latest ChatGPT version, which defaults to Streamable HTTP

Cost considerations

ChatGPT subscription:

• ChatGPT Plus: $20/month - MCP via Developer Mode, individual use
• ChatGPT Pro: $200/month - higher rate limits, priority access
• ChatGPT Team: $25/user/month - workspace sharing, admin controls
• ChatGPT Enterprise/Education: custom pricing - admin-managed Apps, enhanced security

Tallyfy doesn’t charge extra for MCP usage, but all MCP tool calls count toward your Tallyfy API rate limits. Use specific queries to minimize unnecessary tool calls.

Related articles

Mcp Server > Using Tallyfy MCP server with ChatGPT

ChatGPT Enterprise and Team users can connect to Tallyfy’s MCP server at mcp.tallyfy.com through OAuth 2.1 authentication to manage workflows and search tasks and analyze templates using natural language prompts while the text-based interface works well for complex queries and process documentation but falls short for visual workflow tracking and real-time collaboration and bulk operations that are better handled in Tallyfy’s native UI.

Integrations > MCP server

Tallyfy’s MCP Server lets you control workflows through plain English in any major AI platform like ChatGPT or Claude or Gemini or Copilot using the open Model Context Protocol standard to search tasks and launch processes and manage templates and set up automations without needing any API knowledge.

Integrations > BYO AI (Bring Your Own AI)

BYO AI lets you connect your existing AI subscriptions like ChatGPT or Claude directly into Tallyfy workflows through the MCP industry standard so your AI can read task context and complete steps and generate content and make decisions automatically inside running processes without any copy-pasting or app-switching.

Byo Ai > Claude integration

Claude connects to Tallyfy’s MCP server at mcp.tallyfy.com using OAuth 2.1 with PKCE so you can manage tasks and processes and templates and automations through natural language chat in both Claude.ai web and Claude Desktop with over 40 available tools spanning read/write access to every part of your workflow system.

Was this helpful?

YesNo