MCP App Store
by KeenEthics
TODO.work icon

TODO.work MCP Server for ChatGPT

By Wizzo Software ltd

Overview

TODO is a business workspace for tasks, projects, time tracking, team chat, clients and leads. Connect it to ChatGPT to work with your own data: find and create tasks, add comments, move cards on the kanban board, check today’s schedule and time entries, look up clients and colleagues, and ask your workspace knowledge base. Every connection is authorized by the user themselves, acts strictly inside that user’s permissions, can be limited to read-only, and can be disconnected at any time from the workspace settings.

How to connect

  1. ChatGPT: Open the TODO.work listing in the ChatGPT app directory and choose Connect. ChatGPT listing TODO.work exposes 107 tools here.

Tools

add_filter_to_ticket

ChatGPT
Set the value of a custom field on a task: inserts it, or replaces the current value. Get the field ids and their possible values from custom_fields_list.

add_lead

ChatGPT
Create a lead together with its first note in a single call: contact details, source or status, and free text. Prefer it over chaining tickets_create (type lead), add_filter_to_ticket and comments_add. Nothing has to be looked up first: the lead fields are addressed by name, and select fields accept the option's name.

add_project_brief

ChatGPT
Save a project brief: adds the given text to the project as a text item titled 'תיאור הבריף' (brief description) in the project's links. Each call adds another item. (Attaching reference or material FILES to the brief is only possible through the REST multipart API.)

attach_users

ChatGPT
Give one user (usually the client who submitted a project brief) viewer access to a project. By default it also prefixes the project's name with its id, once; pass rename=false to leave the name as it is. Without user_id only the rename happens.

billing_activity_report

ChatGPT
Admin connections only. The activity report (work time and amounts per project) of a client, a user or a date range; at least one of client_id, user_id or date_from with date_to is required.

billing_activity_report_pdf

ChatGPT
Admin connections only. The same activity report as a PDF file, the file the reports screen downloads. Returns {file_name, mime, size, data_base64}: the SAME PDF the reports screen downloads, ready to attach via chat send_file

billing_add_note

ChatGPT
Admin connections only. Append a dated line to a plan row's comments (the collection diary) without touching its other fields. Append a dated line to the row's comments without touching other fields (the collection diary).

billing_client_debt

ChatGPT
Admin connections only. The unpaid plan rows of one client up to a date.

billing_debts

ChatGPT
Admin connections only. The unpaid amounts per client up to a date.

billing_issue_document

ChatGPT
Admin connections only. Issues a REAL document at the invoice provider for a plan row: a payment request (type 300, the default) or a tax invoice-receipt (type 320, which also marks the row paid). It cannot be undone; only on the user's explicit instruction. Create_doc issues a REAL document at the invoice provider.

billing_plan

ChatGPT
Admin connections only. The collection plan rows of a month (default: the current one) plus unpaid rows carried over from earlier months. Amounts are stored before VAT; a row with vat=1 is charged amount x (1 + VAT).

billing_save_row

ChatGPT
Admin connections only. Add a row to the collection plan, or update a row when id is given (the fields you pass replace the current values). Serves a reconciliation the user approved. Note: add{status:'paid'} stamps paid_date as the 1st of billing_date; to mark paid with the REAL payment date issue the receipt instead

billing_send_payment_request

ChatGPT
Admin connections only. Emails the client the payment-request link of a plan row (issuing the document first when there is none) and writes the collection diary. A REAL email to the client; only on the user's explicit instruction. Emails the חשבון עסקה link (issuing it first if none) and writes the diary line; a REAL email to the client

bots_list

ChatGPT
Admin connections only. The workspace's bots: bot id, user id, name and active scopes. Returns: the platform's bots (bot_id, user_id, name, active scopes)

bots_remove_kringl

ChatGPT
Admin connections only. Retires the Kringl bot user: its webhook endpoint is deleted, the user is removed from the team and every API key bound to it stops working. Returns: {removed, bot_user_id}.

bots_routines

ChatGPT
Admin connections only. A bot's scheduled routines with their last and next run. Returns: the bot's scheduled routines with their last and next run.

bots_run

ChatGPT
Admin connections only. One run of a bot with its full list of steps. Returns: one run with its full step list

bots_runs

ChatGPT
Admin connections only. A bot's run log, newest first: what triggered each turn, which tools it called, how long it took, its token cost and what failed. Returns: newest first

bots_stats

ChatGPT
Admin connections only. A bot's runs per day and per trigger and its tool-call counts. Returns: {by_day[], by_trigger[], tools[{tool, calls, failed}]}

call_transcripts

ChatGPT
Transcripts of the voice calls a user took part in.

chat_add_participants

ChatGPT
Add users to a group conversation.

chat_conversations

ChatGPT
List the acting user's chat conversations, private chats and groups, newest first.

chat_create_group

ChatGPT
Create a group conversation with the given participants.

chat_leave_group

ChatGPT
The acting user leaves a group conversation.

chat_mark_read

ChatGPT
Mark every message of a conversation as read by the acting user.

chat_messages

ChatGPT
Read the messages of a conversation, newest first; page back with before_id.

chat_open_private

ChatGPT
Get the private conversation with a user, creating it when it does not exist yet.

chat_remove_participant

ChatGPT
Remove a participant from a group conversation.

chat_rename_group

ChatGPT
Rename a group conversation (replaces its name).

chat_send_file

ChatGPT
Send a file, given as a data URI, to a conversation, optionally with a caption. Everyone in it is notified.

chat_send_message

ChatGPT
Send a message to a conversation. Everyone in it is notified, and a conversation may include the workspace's clients. Structured blocks (facts, tables, lists, charts) can ride under the text.

chat_set_admin

ChatGPT
Make a participant an admin of a group conversation, or take the admin rights away.

chat_users

ChatGPT
List the users the acting user can start a conversation with.

comments_add

ChatGPT
Add a comment on a ticket (supports mentions, replies and internal notes). visibility 'internal' = a staff-only note: hidden from client (external) users everywhere — ticket view, list previews, search, notifications and Ask Todo. audience narrows WHO on the staff side sees it: 'staff' (default, everyone on the team), 'handlers' (the task handlers, the managers above them in the org tree, and admins) or 'admins' (system admins only). A reply always inherits its parent's visibility and audience. Only staff users and bots may write internal notes; a bot key without an explicit visibility follows the bot's 'comment visibility' default from the bots screen.

custom_fields_add_to_project

ChatGPT
Make an existing custom field show in one more project. Workspace admin only. Returns: {id, project_id, filter}

custom_fields_for_target

ChatGPT
Custom field definitions of one record type: 'task', 'lead' or 'expense'.

custom_fields_list

ChatGPT
Definitions of the workspace's custom fields: id, name, type and the label of every possible value. Needed to read a task's custom_filters map and to set a value with add_filter_to_ticket. Pass no parameters to get every field. Pass no params to get EVERY field (a project_id/target narrows the list and may hide project-specific fields)

custom_fields_remove_from_project

ChatGPT
Stop showing a custom field in one project. The field and the values tasks hold stay. Workspace admin only. Returns: {id, project_id, filter}

custom_fields_save

ChatGPT
Create a custom field (a column on tasks, leads or expenses) or update an existing one by id: its name, its type, the options of a select field, the record types it applies to, the projects it shows in. Workspace admin only. Never deletes: on an update the options you send are added to the existing ones. Call custom_fields_list first, a field with that name may already exist. Returns: {id, created, filter}; then set a value on a task with add_filter_to_ticket (a dropdown field takes the option's id from filter.values)

daily_target_progress

ChatGPT
Read-only: how far one user has got toward their daily target on a given day (default today): the time logged on task timers against the daily target set on their user card (hours, or another target type), the same numbers as the target clock on the dashboard. A daily_target of 0 means no target is set.

expenses_applicable_rules

ChatGPT
Which approval rules apply to an amount / supplier / budget / category (or to a saved expense), the fields they require, and who will approve. Returns: {rules[{id, name, approver_user_id, approver_name, fields[]}], fields[{id, label, type('text'|'date'|'file'|...), required}], missing_required_fields[], rule_approvers[], budget_default_approvers[]}

expenses_approve

ChatGPT
Approve an expense as its approver (the acting user's own approval; an expenses manager may act on another approver's). Fails with the list of missing required rule fields when the approval rules demand them. / reject {id*, notes?} → {success, expense_status, approvals[]}: sets the acting user's own approval on the expense; the expense turns approved when every approver approved, rejected when one rejected; error_code 'not_an_approver' | 'no_approvers' | 'missing_rule_fields'(with missing_fields[])

expenses_attach_document

ChatGPT
Attach an already uploaded document (storage_id) to an expense.

expenses_budget

ChatGPT
One budget with its default approver(s) and members. Returns: one budget, same shape

expenses_budgets

ChatGPT
The budgets the acting user sees, each with its default approver(s), members and default category. Returns: [{id, name, monthly_budget, default_category_id, default_approvers[{user_id, name}], members[{user_id, name, can_add_expense, can_edit_expenses, can_view_all, is_default_approver}]}]

expenses_categories

ChatGPT
The expense categories (id and name). Returns: [{id, name}]

expenses_create_supplier

ChatGPT
Create a supplier with a name and, optionally, a business number and contact details.

expenses_delete

ChatGPT
Move an expense to the archive. The row goes to the archive (is_deleted), refused for a paid or locked row the same way the screen refuses

expenses_exchange_rate

ChatGPT
The Bank of Israel rate of a currency for a day (on weekends and holidays, the last rate published before it). Returns: {rate, rate_date(the day actually published: Fri/Sat/holidays get the last one before), requested_date, source('BOI'|'ECB'|'FIXED')}

expenses_export_documents

ChatGPT
Build a ZIP of the expense documents of a period and return its download link. Returns: {download_url, filename, file_count}, the same ZIP the reports screen downloads.

expenses_file_document

ChatGPT
Read a supplier invoice or receipt (PDF or image) with the AI document reader and file it: attach it to the matching expense, or open a new pending expense when none matches. Returns: {action: 'attached'|'created'|'duplicate'|'needs_decision'|'no_match'|'already_attached'|'own_document'|'not_invoice'(a quote, a report, a statement, a screenshot: not a billing document)|'third_party_document', expense_id?, expense?, pending_approval?(created rows: true when it waits for the approver), match{storage_id, extracted{..., customer_name, document_kind('invoice'|'receipt'|'invoice_receipt'|'credit_note'|'proforma'|'quote'|'report'|'statement'|'other')}, supplier, similar_expense}, note?}; a document with no readable supplier is 'needs_decision' (reason no_supplier), never a row with an empty supplier

expenses_find_similar

ChatGPT
Find an expense of the same supplier and amount in the same month, before filing a duplicate.

expenses_gaps

ChatGPT
The gaps of a month: expenses without a document, recurring templates that produced no expense, and unpaid expenses. Returns: {missing_document[], templates_without_row[], unpaid[], require_document}

expenses_get

ChatGPT
Read one expense with its documents.

expenses_mark_paid

ChatGPT
Mark an expense as paid, or toggle it back.

expenses_recurring

ChatGPT
List the recurring expense templates.

expenses_reject

ChatGPT
Reject an expense as its approver, with an optional note.

expenses_remove_document

ChatGPT
Take a document off an expense. Takes the document off the row (the file is deleted only when no other row holds it)

expenses_rules

ChatGPT
The workspace's expense approval rules (amount tiers, required fields, extra approver). Expenses managers only. Returns: the approval rules as the settings screen keeps them (expenses managers only)

expenses_save

ChatGPT
Create an expense, or update one when id is given (the fields you pass replace the current values). Documents are NOT changed by save: use attach_document / remove_document

expenses_suppliers

ChatGPT
Search the workspace's suppliers.

expenses_upload_file

ChatGPT
Store a file and get its storage_id, for a file field of an approval rule or for attach_document. Returns: {storage_id}

get_clients

ChatGPT
List the workspace's clients.

get_project_senior_members

ChatGPT
Top-permission team members of a project (falls back to platform admins).

get_users

ChatGPT
List the workspace's users that the acting user may see: id, display name and type. The row with is_me = true is the acting user, but prefer whoami for that.

kanban_board_tasks

ChatGPT
The task cards of a kanban board, list by list.

kanban_boards

ChatGPT
List or search the kanban boards the acting user can see.

kanban_create_board

ChatGPT
Create a kanban board, optionally shared with a user.

kanban_delete_board

ChatGPT
Delete a kanban board.

kanban_delete_list

ChatGPT
Delete a list (column) from a kanban board.

kanban_find_shared_board

ChatGPT
Find the kanban board two users share. Admin connections only.

kanban_lists

ChatGPT
List the lists (columns) of a kanban board.

kanban_move_task

ChatGPT
Place a task's card in a list, or move it there from another list of the board.

kanban_remove_task

ChatGPT
Take a task's card off a list; the task itself stays.

kanban_save_list

ChatGPT
Create a list (column) on a kanban board, or rename an existing one when id is given.

meetings_cancel

ChatGPT
Cancel a meeting the acting user organizes; every attendee gets a cancellation email. Returns: {status:'cancelled', attendees_notified}, organizer only

meetings_create

ChatGPT
Create a calendar meeting organized by the acting user. Its attendees, people outside the workspace included, get an invitation email with a calendar file. Returns: the meeting + notified

meetings_free_slots

ChatGPT
Find the time windows in which all the given users are free. Returns: {days[{date, weekday, free[{start, end, start_at, end_at}], busy{user_id: [[start, end]]}}]}: the windows of at least duration_min in which ALL of them are free; busy counts meetings, appointment bookings, their synced Google Calendar and reported absences, and nothing earlier than now is offered.

meetings_get

ChatGPT
Read one calendar meeting. Returns: the meeting with start_at/end_at, status ('confirmed'|'cancelled'), attendees with rsvp, reminders, join_url

meetings_invitees

ChatGPT
Search the people a meeting can invite. Returns: {users[{user_id, name, email, is_team}]}, everyone a meeting can invite

meetings_update

ChatGPT
Change a meeting the acting user organizes (subject, time, place, attendees, reminders); the values you pass replace the current ones and the attendees get an update email. Organizer only

projects_archive

ChatGPT
Archive (status true) or restore (status false) a project. Reversible; the billing history is kept. Reversible, keeps the billing history

projects_delete

ChatGPT
Permanently delete an archived project that has no logged work time, together with its tasks and files. Irreversible: archive instead unless the user explicitly asks to destroy it. PERMANENT and irreversible: erases the project, its tasks and its files. Only accepts an already-archived project with NO logged work time on any of its tasks; otherwise it refuses with 400. Archive instead unless the caller really means destroy.

projects_folders

ChatGPT
List the project folders (id, name, color) that a project's folder_id refers to. (resolves the folder_id every project carries → {id, name, color})

projects_get

ChatGPT
Read one project with its full details.

projects_save

ChatGPT
Create a project, or update an existing one when id is given (the fields you pass replace the current values).

tickets_add_handler

ChatGPT
Add a user as a handler (assignee) of a task.

tickets_create

ChatGPT
Create a task in a project (or a lead, with type 'lead'), optionally assigned to handlers, scheduled, or placed in a kanban list.

tickets_create_subtask

ChatGPT
Create a subtask under an existing task.

tickets_edit_time

ChatGPT
Change the minutes, and optionally the date or description, of a closed time entry. Replaces the previous values.

tickets_get

ChatGPT
Read one task with its details: subject, status, project, handlers, dates, custom field values (custom_filters) and logged time.

tickets_log_time

ChatGPT
Add a work-time entry, in minutes, on a task for the acting user or another handler, optionally for a past date and with a short description of what was done.

tickets_remove_handler

ChatGPT
Remove a user from a task's handlers.

tickets_set_time_description

ChatGPT
Set the description of a logged time entry (by its id, or the acting user's latest entry on a task). Replaces the existing description.

tickets_start_timer

ChatGPT
Start the acting user's live work timer on a task.

tickets_stop_timer

ChatGPT
Stop the acting user's running work timer on a task; the elapsed time is logged.

tickets_update

ChatGPT
Change one field of a task (for example subject, status, scheduled date or a custom field). The field's previous value is replaced.

tracking_active_shifts

ChatGPT
Workspace-wide: the users whose attendance shift is open right now. Needs the time-tracking permission.

tracking_day_timers

ChatGPT
Workspace-wide: the task timers closed today, or on a given date. Needs the time-tracking permission.

tracking_running_timers

ChatGPT
Workspace-wide: the task timers running right now, who works on what. Needs the time-tracking permission.

whoami

ChatGPT
Who this connection acts as: the user every other call runs as (id, name, admin or not, workspace). Call it before anything phrased as "me" or "my"; do not guess the acting user from get_users.

Capabilities

Writes

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/todo-work"><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/todo-work)