add_activities
ChatGPTCreates multiple activities in batch. Required: activities list with name for each. Optional per activity: code, color (HEX only; agent converts names to hex), description, and assignedGroups (array of groupId GUIDs resolved by the agent via get_all_groups if needed). Returns list of created activities with activityId, activityName, and activityCode. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
add_clients
ChatGPTCreates multiple clients in batch. Required: clients list with name for each. Optional per client: code, projects (array of project ID GUIDs to associate with the client, resolved by the agent via get_all_projects if needed). Returns list of created clients with clientId and clientName. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
add_hour_entries
ChatGPTCreates one or more manual Hour Entries for a person with the specified parameters. This allows recording worked hours directly instead of live clock in/out. To find the most suitable project/activity/location, use 'get_all_activities', 'get_all_projects', and 'get_all_locations'. If multiple locations share the same name, prompt the user to select the correct location. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
add_projects
ChatGPTCreates multiple projects in batch. Required: projects list with name for each. Optional per project: code, description, clientId (client GUID resolved by the agent via get_all_clients if needed), assignedGroups (array of groupId GUIDs resolved by the agent via get_all_groups if needed), and locationId (location GUID resolved by the agent via get_all_locations if needed). Returns list of created projects with projectId, projectName, and projectCode. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
add_schedule
ChatGPTCreates a new schedule with specified work arrangements and day schedules. For Weekly arrangements, weeklyHours is required. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
assign_person_to_position
ChatGPTAssigns a position to a person. If positionId is omitted (null), Guid.Empty, or '11111111-1111-1111-1111-111111111111', the person's position is cleared. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
assign_schedules
ChatGPTAssigns one or more schedules to a list of people and/or groups. Use get_all_schedules to get schedule IDs, get_managed_people to resolve person names to IDs, and get_all_groups to resolve group names to IDs. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
clock_in
ChatGPTCreates Clock In TE for person with given parameters. Past timestamps are allowed when provided. Future scheduling is not supported (except auto clock-out set during clock-in). To find the most suitable project/activity/location, use 'get_all_activities', 'get_all_projects', and 'get_all_locations' tools. If multiple locations share the same name, prompt the user to select the specific location to clock in to. MANDATORY: Always call this tool when a user requests to clock in, regardless of previous conversation history or perceived status. This tool is the ONLY source of truth for the user's current clock-in state. If the user is already clocked in, the backend will return an error (if any), which you must then report. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
clock_out
ChatGPTCreates Clock out TE for person with given parameters. Past timestamps are allowed when provided. Future scheduling is not supported. To find the most suitable location, use 'get_all_locations' tool. If multiple locations share the same name, prompt the user to select the specific one to clock out from. MANDATORY: Always call this tool when a user requests to clock out, regardless of previous conversation history or perceived status. This tool is the ONLY source of truth for the user's current clock-out state. If the user is already clocked out, the backend will return an error (if any), which you must then report. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
delete_schedules
ChatGPTDeletes a list of schedules by their IDs. Use get_all_schedules to get schedule IDs if not known. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
end_break
ChatGPTEnds an in-progress break for a person. The backend creates a clock-in (In) TE that resumes work, automatically restoring the project and activity from the clock-in entry before the break. Completes the flow started by 'start_break'. Supports ending a break at a specific past time using the time parameter. Future scheduling is not supported. To find the most suitable location, use 'get_all_locations' tool. To see who is currently on break, use 'get_who_on_break' tool. MANDATORY: Always call this tool when a user requests to end a break or resume work after a break, regardless of previous conversation history or perceived status. This tool is the ONLY source of truth for the user's current break state. If the person is not on break, the backend returns an error (e.g., 'Invalid operation, nothing to end'), which you must then report. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_all_activities
ChatGPTReturns a paginated list of activities (activityName, activityId, activityCode) available for person. By default returns only Active activities. Use status parameter to filter by specific status or get all. Use code parameter to filter by activity code. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_all_clients
ChatGPTReturns a paginated list of clients (clientId, clientName, clientCode) available for the person. By default returns only Active clients. Use status parameter to filter by specific status or get all. Use code parameter to filter by client code. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_all_groups
ChatGPTReturns a paginated list of groups (groupId, groupName) available to a person (all groups from organization, managed group for manager). By default returns only Active groups. Use status parameter to filter by specific status or get all. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_all_locations
ChatGPTReturns a paginated list of locations (locationName, locationId), available for person. By default returns only Active locations. Use status parameter to filter by specific status or get all. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_all_positions
ChatGPTReturns a paginated list of positions (positionId, positionName) available for the person. By default returns only Active positions. Use status parameter to filter by specific status or get all. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_all_projects
ChatGPTReturns a paginated list of projects (projectName, projectId, projectCode) available for person. By default returns only Active projects. Use status parameter to filter by specific status or get all. Use code parameter to filter by project code. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_all_schedules
ChatGPTReturns a paginated list of schedules (scheduleName, scheduleId, workArrangement, weeklyHours, daySchedules) available for person. For Fixed arrangements, daySchedules carries from/to per day. For Flexible arrangements, daySchedules carries duration per day. For Weekly arrangements, weeklyHours is set and daySchedules is null. Optionally filter by specific person, group, or work arrangement type (Fixed, Flexible, Weekly). Use get_managed_people to resolve person names to IDs, or get_all_groups to resolve group names to IDs. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_all_time_off_policies
ChatGPTReturns a paginated list of time-off policies. By default returns only Active policies. Use status parameter to filter by specific status or get all. If the user asks about a policy's details (e.g., accrual rules, compensation type, balance rules), the agent SHOULD use the detailed description of each property in the Policy and AccrualSettings models. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_api_docs
ChatGPTUse this tool when the user asks a question related to the Jibble API. It returns a list of API documentation that provides technical context, guidance, and reference information about Jibble's API endpoints. IMPORTANT for multilingual users: If the user asks in another language, translate the API question to English before calling this tool, then explain the answer back in the user's original language. This tool is ideal for retrieving relevant API documentation and technical specifications that can help the user understand how to integrate with and use the Jibble API effectively. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_app_ui_navigations
ChatGPTUse when users ask how to navigate to/find Jibble pages or features. Returns role-filtered routes with URLs. For multilingual queries: infer/translate the navigation intent to English when matching feature names and route keywords, but keep URLs unchanged and reply to the user in their original language. CRITICAL: Only provide complete, functional URLs with all data filled in. NEVER show templates, placeholders (:id, {param}, YYYY-MM-DD), or URL patterns. If parameters missing: State what's needed, wait for data, then show complete URL only. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_attendance_insights
ChatGPTReturns punctuality and attendance analytics for specified people within a date range: attendance score (0-100, percent of scheduled work days the person showed up), counts of on-time / late / no-show / rest days, average first clock-in time, average last clock-out time, and average break duration. Use this tool when the user asks about punctuality, lateness, absenteeism, no-shows, attendance scores, or average start/end times. 'PersonIds' is optional: leave it out to cover everyone the caller can see (admins/owners: the whole organization; managers: their managed people). Without 'GroupBy' that is one page of people per call, so use 'Page' to walk through a large workspace; with 'GroupBy' the whole visible population is aggregated at once and the call fails with guidance when it exceeds 150 people. If the user asked about specific people, resolve the person IDs via 'get_managed_people' first; for a whole team/group, resolve member IDs via 'get_group_persons' first and pass them all. At most 150 person IDs can be passed in one call. Set 'GroupBy' to 'Group', 'Position', or 'Location' to get aggregated rows per group/position/location (each with its member breakdown in 'Members'); leave it unset (or 'None') for one row per person. Dates use yyyy-MM-dd format (e.g., '2026-06-01'); both StartDate and EndDate are inclusive, and the period cannot cover more than 30 calendar dates (e.g. 2026-06-01 to 2026-06-30) — split a longer question into consecutive ranges of at most 30 dates rather than asking for one wide range. Times returned (average clock in/out) are in each person's own timezone context as recorded by the backend. Set 'IncludeDailyStatuses' to true only when the user asks which specific days a person was absent/late — it adds a per-day status list and makes the response much larger. NOTE: late-arrival detection depends on assigned schedules; people without schedule data are only classified as present or no-show. For worked-hours totals use 'get_timesheets'; for live presence use 'get_team_attendance_dashboard'. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_clock_in_patterns
ChatGPTReturns a clock-in pattern analysis for a group over a date range. Shows each person's average first clock-in time, earliest, latest, and number of days tracked. Helps identify punctuality patterns. GroupId is required for scalability. Use for queries like 'Show clock-in patterns', 'Who arrives earliest?', 'Clock-in times for my team this month'. Max 90-day range. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_geofence_compliance_dashboard
ChatGPTReturns a geofence compliance dashboard for a person or group over a date range. Reports clock-in/out entries that the backend flagged as outside the assigned geofence. Counts flagged and not-flagged entries. The flag does not establish GPS availability or actual compliance: entries without GPS can be unflagged. Date boundaries are UTC; occurrence times retain recorded offsets. Provide either PersonId or GroupId. When PersonId is provided, it is used directly and GroupId is ignored. Date range defaults to the last 30 days when omitted. Use for queries like 'Who clocked in outside their geofence?', 'Show geofence violations this week', 'Geofence compliance for the sales team'. Max 30-day range to keep query cost bounded for large groups. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_group_persons
ChatGPTReturns a paginated list of people (personId, personName, positionId, positionName, email, phoneNumber, code, role, status) from a specific group, along with the total count. Supports search by fullName, email, phoneNumber, code, positionId. If user provides a position name, first use get_all_positions to resolve the matching positionId. Use filterMode='Any' (default, OR) or 'All' (AND) to combine multiple search fields. To find group by group name, use 'get_all_groups' tool This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_group_settings
ChatGPTReturns list of all settings used for specific group. Use get_settings_info tool to retrieve settings names, descriptions and value types. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_leave_balance
ChatGPTGets leave balance as of a specific date, with optional filtering by leave policy and person. Date parameter is REQUIRED and must be explicitly provided by the user. If no date is provided, inform the user that a specific date is required. Use policyName when user mentions specific leave types. For person names, resolve ID via get_managed_people first. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_leave_history
ChatGPTReturns a person's comprehensive time-off history in a requested date range to EXPLAIN balance differences, policy assignments, administrative actions, and leave-related changes. The AGENT SHOULD: fetch entries; analyze by date range, CalculationType, and administrative actions; identify who performed actions and when; sum changes by type; and provide detailed explanations for balance discrepancies, policy changes, and administrative history. BOTH START DATE AND END DATE ARE REQUIRED. Use it to explain one person's leave balance changes; do not use it to assemble a team report. For everyone's current balances, call get_leave_balance without a person. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_leave_utilization
ChatGPTReturns a leave utilization dashboard showing how much of their entitled leave each person has used. Helps identify people with excess balance or high usage. Use for queries like 'Show leave utilization', 'Who has unused leave?', 'Team leave balance overview'. Optional GroupIds to filter by team. Optional Date for as-of balance (defaults to today). This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_managed_people
ChatGPTReturns a paginated list of people (personName, personId, positionId, positionName, email, phoneNumber, code, role, status) that the current person can act on behalf of, along with the total count. Supports search by fullName, email, phoneNumber, code, positionId. If user provides a position name, first use get_all_positions to resolve the matching positionId. Use filterMode='Any' (default, OR) or 'All' (AND) to combine multiple search fields. When any search filter is provided, returns people regardless of status. For managers - manageable via group or directly. For Admins/Owners - all people. For others - only themselves This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_my_work_summary
ChatGPTReturns a work summary for the authenticated user. Shows hours worked this week, expected hours, days worked, on-track status, a daily hours bar chart over the lookback period, and leave balance per policy. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_onboarding_status
ChatGPTShows the organization's Jibble setup progress as an interactive checklist — which onboarding steps (work schedule, time tracking rules, activities & projects, locations, team members) are done vs remaining, plus overall progress. Mirrors the in-app Get Started experience. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_organization_settings
ChatGPTReturns list of all settings and exact values used in the organization. Use get_settings_info tool to retrieve full settings list with settings names, descriptions and value types. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_overtime_analysis
ChatGPTReturns an overtime analysis dashboard showing breakdown across all overtime categories. Categories: Daily, Double, Rest Day, Public Holiday, Weekly. Shows which people have the most overtime and which types are most common. Use for queries like 'Show overtime breakdown', 'Who has the most overtime?', 'Overtime analysis for my team this month'. Requires exact start and end dates (yyyy-MM-dd). Optional GroupIds to filter. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_people_by_schedule
ChatGPTReturns a paginated list of people (personId, personName, email, phoneNumber, code, role, status, timeSheetTimeZone) assigned to a specific schedule. Use 'get_all_schedules' to find schedule IDs by name or work arrangement type. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 250). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_person_assignments
ChatGPTReturns everything a person can track time against, in one call: assigned activities, assigned projects, effective work schedule, upcoming holidays, and geofenced locations (allowed clock-in areas). Use this before clock_in or add_hour_entries to resolve valid activity/project IDs for a person, or to answer questions like 'what can I clock in to?', 'what is my schedule?', 'where am I allowed to clock in?'. If otherPersonId is omitted, returns assignments for the current person. If only a person name was provided, resolve the GUID via get_managed_people first. Sections that fail to load are reported in sectionErrors without failing the whole call. For full break/overtime/auto-deduction details use get_user_settings; for org-wide lists use get_all_activities / get_all_projects. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_person_work_summary
ChatGPTReturns a work summary for a specified person. Only available to Owner and Admin roles. If the user provided a person name, resolve the person ID via get_managed_people first. Shows hours worked this week, expected hours, days worked, on-track status, a daily hours bar chart over the lookback period, and leave balance per policy. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_public_holidays
ChatGPTGets a calendar (e.g., public holidays). If otherPersonId is provided, filters by that person. If only a person name is provided, resolve it via get_managed_people; otherwise, use the current person. When no holiday calendar is configured, returns a successful empty result with IsConfigured=false; do not retry unless calendar configuration changes. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_settings_info
ChatGPTReturns a list of all settings can be configured for Organization in Jibble with detailed information about each setting. Use this tool to provide more detailed information about settings in additional to the information from help articles. This tool can be used to answer user's questions related to settings, such as: 'Which settings can be configured for Organization?', 'Which settings are configurable for Organization?', 'What is the default value/possible values for setting X?'. Also it can be used to build an input for edit_settings tool. Each setting property includes 'restrictions' indicating which subscription plans restrict that setting and how. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_subscription_feature_catalogue
ChatGPTReturns the calling organization's current subscription plan, its effective feature restrictions and limits (including any organization-level overrides), and the full catalogue of subscription tiers with the restrictions imposed by each tier. Use this tool for any question about plans, subscriptions, features, feature availability, or feature limits/quotas. Call it BEFORE searching help articles for these topics — this tool is authoritative for the caller's actual plan and limits, whereas help articles describe the product generically and do NOT reflect the caller's actual entitlements. Feature identity for comparison/explanation is the (Code, Type) pair on each restriction row — use these stable codes when explaining which tier unlocks a feature. For entitlement actions such as starting a feature trial, use the restriction's FeatureIds (the product-feature GUIDs) as the feature identifier — never the Code or display name. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_team_attendance_dashboard
ChatGPTReturns backend-calculated worked time for the selected timesheet date, with separate live clock status for today (UTC). Zero worked time is not an absence determination: schedules, holidays and approved leave affect absence. Requires a group ID for scalability. Defaults to today (UTC) if no date is provided. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_team_schedule_view
ChatGPTReturns current schedule assignments and backend-calculated worked time for the selected timesheet date. Zero worked time does not imply absence. Schedule names are current assignments, not proof of being scheduled on the selected date. Live clock status is included separately only for today (UTC). GroupId is required for scalability — no org-wide scans. Defaults to today (UTC) if no date is provided. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_team_time_insights
ChatGPTReturns an interactive time distribution dashboard grouped by Person, Location, or Group. Use for queries like 'Show hours by location', 'Time spent per group/center', 'Who worked the most this month', 'Hours distribution by team'. GroupBy=Person: hours per person. GroupBy=Group: hours per team/center. GroupBy=Location: each person's full-period worked hours allocated to their most frequent active clock-in location, using UTC dates for clock-ins; not actual duration worked at each location. Requires exact start and end dates (yyyy-MM-dd). Optional GroupIds to filter. To compare group totals over one interval, call once with GroupBy=Group (paginated) instead of once per group. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_time_entries
ChatGPTReturns raw clock-in, clock-out, and break records, newest first. Omit person selectors for yourself, use OtherPersonId for one person, or PersonIds for up to 100 people within your permitted scope. PersonIds requires StartDate and EndDate (yyyy-MM-dd), inclusive UTC days, covering at most 31 days inclusive; the dates also work with a single person. They are UTC days, not timesheet-local: for a local calendar day, widen the range by one day on each side and keep the entries whose timestamp (local time) falls on that day. Each page covers the entire selected population; read pages sequentially. For worked-hours totals or daily summaries, use get_timesheets instead of calling this once per person. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. Maximum 100 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
get_time_off_requests
ChatGPTReturns raw time-off request records (dates, status, duration, policy) for a period as paginated JSON. Owners and admins asking to see, list, show or review time-off requests for their team or a period, with any status, should use get_time_off_requests_view instead: it returns the same requests as an interactive queue with status counts and approve and reject controls. Use this tool when the caller is a member or manager (they cannot call the view), when the request is about one person's own requests, or when the raw records are needed for a calculation or export rather than for display. If the user provided person names, resolve the person IDs via get_managed_people first. For detailed leave history including balance changes, policy assignments, administrative actions, and explanations of balance discrepancies, use the get_leave_history tool instead. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_time_off_requests_for_groups
ChatGPTReturns a paginated list of all time-offs requested by persons from specified groups for specified period This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_time_off_requests_view
ChatGPTShows the team's time-off requests for a period as an interactive queue without modifying any requests. This is the default tool whenever an owner or admin asks to see, list, show or review time-off or leave requests — all statuses, pending only, approved and rejected included, for the whole team, a group or a person — not just when approval is mentioned. Prefer it over get_time_off_requests, which returns the same records as plain JSON. Includes status counts and a paginated table with approval controls; opening, filtering and paging only read data. Optional GroupIds, PersonIds and Status filter the queue. Dates default to the current calendar month. Owners and admins only; members and managers use get_time_off_requests. Only an explicit decision uses manage_time_off_requests; viewing this widget does not approve or reject anything. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
get_time_tracking_report
ChatGPTRetrieves time tracking data with flexible grouping and filtering. Use this tool when the user asks about hours spent, time allocation, usage patterns, or comparisons for projects, clients, or activities. This includes queries like: 'How many hours for Project X?', 'Most/least used project', 'Top N projects/activities/clients by hours', 'Projects with no tracked time', 'Highest billable amount', 'Compare Project A vs B', or 'Resource allocation analysis'. IMPORTANT: The tool requires specific start and end dates (yyyy-MM-dd format). When users mention relative periods like 'last week', 'this month', 'this quarter', or 'all time', you MUST ask them to provide exact calendar dates before calling this tool. Set GroupBy to 'Project' for project analysis, 'Client' for client analysis, or 'Activity' for activity-based reports. Use ProjectIds, ClientIds, or ActivityIds to filter specific entities (resolve names to GUIDs using get_all_projects, get_all_clients, or get_all_activities first). Returns aggregated time entries grouped by the specified dimension with detailed breakdowns including total hours, daily items, and billable information. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_timesheets
ChatGPTReturns timesheets summary for specified persons within a custom date range. Use this as the default method for getting timesheets. You can optionally filter by specific person IDs. If the user provided person names, resolve the person IDs via get_managed_people first. Result will include summary timesheet for each person and a set of daily timesheets. Days with no tracked time are excluded from the response. Fields with zero values are omitted from the response. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_timesheets_for_groups
ChatGPTReturns timesheets summary for ALL persons from specified groups within a custom date range. Use this method ONLY when the user specifically wants to filter by groups/teams. Result will include summary timesheet for each person and a set of daily timesheets. Days with no tracked time are excluded from the response. Fields with zero values are omitted from the response. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. This tool does not support batching for write operations. Repeated calls are allowed only for pagination/reading pages; do NOT use multiple calls to bypass operation limits for a single user request.
get_user_settings
ChatGPTReturns all settings affecting clock-in, clock-out, and break flows. If anotherPersonId is provided, fetch settings for that person. If only a name was supplied, resolve the GUID via get_managed_people. If neither is given, return the settings for the current person. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_user_time_sheet_time_zone
ChatGPTReturns the timesheet time zone for a person. If anotherPersonId is provided, fetch it for that person. If only a name was supplied, resolve the GUID via get_managed_people. If neither is given, return the time zone for the current person. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_who_clocked_in
ChatGPTReturns a paginated list of people who are currently clocked in. Optionally filter by a list of person IDs to check clock-in status for specific people. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. Maximum 50 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
get_who_clocked_out
ChatGPTReturns a paginated list of people who are currently clocked out. Optionally filter by a list of person IDs to check clock-out status for specific people. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. Maximum 50 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
get_who_on_break
ChatGPTReturns a paginated list of people who are currently on break. Optionally filter by a list of person IDs to check break status for specific people. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. Maximum 50 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
get_wizard
ChatGPTReturns the full wizard definition with all steps for a given wizard ID. Use this tool after calling 'get_wizards' to retrieve the complete step-by-step instructions for a specific wizard. Follow the steps in order to answer the user's question. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
get_wizards
ChatGPTReturns a list of available guided workflows (wizards) that help answer complex questions requiring multiple steps. Use this tool when the user asks a complex or unclear question that may require calling multiple tools in sequence, analyzing intermediate results, or combining data from different sources. Each wizard provides a structured, step-by-step plan for the AI to follow. After finding a relevant wizard, use 'get_wizard' to retrieve the full step-by-step instructions. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
manage_activities
ChatGPTManage activities by updating, archiving, or unarchiving them in batch. Use get_all_activities to retrieve activity IDs. To clear optional fields: use '__CLEAR__' for string. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
manage_activity_group_assignment
ChatGPTManage group assignments for activities. Use get_all_activities to retrieve activity IDs and get_all_groups to retrieve group IDs. Can add groups, remove groups, or replace all groups for an activity. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
manage_groups
ChatGPTManages groups in batch: create, rename, archive/unarchive groups, move people in or out of groups, and set group managers and pay-period approvers (first/second level). Provide at least one of the operation lists. Operations are processed in this order: addGroups, renames, archiveGroupIds, unarchiveGroupIds, memberAssignments, setManagers, setFirstApprovers, setSecondApprovers. Use get_all_groups to resolve group names to IDs and get_managed_people to resolve person names to IDs. A person can belong to only ONE group — assigning a member to a group moves them out of their previous group. Use groupId '11111111-1111-1111-1111-111111111111' in memberAssignments to remove a person from their group. Managers and approvers are replaced as a whole set per group: by default assignPersonIds are merged with the existing set and unassignPersonIds removed; set replaceAll=true to overwrite the set with exactly assignPersonIds. Archiving a group does not archive its members. New groups inherit organization-level settings automatically. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
manage_people
ChatGPTManages workspace members in batch: invite new people, update profiles (name, email, phone, code, group, position, employment type, work start date), change roles, archive or restore members, resend pending invitations, and reset kiosk PINs. Provide at least one of the operation lists. Operations are processed in this order: invites, updates, setRoles, archivePersonIds, restorePersonIds, resendInvitationPersonIds, resetPins. Use get_managed_people to resolve person names to IDs, get_all_groups for group IDs, and get_all_positions for position IDs. Role changes accept Member|Manager|Admin only — ownership transfer is NOT supported and must be done in the Jibble web app. Email and phone number of a person who has already accepted their invite CANNOT be changed (only that person can change them in their own account). Archiving sets the member status to Removed (the Jibble UI shows this as an archived member); it does not delete any data and can be undone with restorePersonIds. To clear optional string fields in updates use '__CLEAR__'; to clear group/position/employment type use '11111111-1111-1111-1111-111111111111'. MANDATORY: before archiving or resetting PINs, confirm the exact person with the user if multiple people match a provided name. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
manage_positions
ChatGPTCreates and/or updates positions in batch. Provide addPositions to create, updates to rename existing positions, or both. Returns succeeded positions and any per-item errors when part of the batch fails. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
manage_project_client_assignment
ChatGPTManage client assignment for a project. Use get_all_projects to retrieve project IDs and get_all_clients to retrieve client IDs. Can assign or clear client for a project. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
manage_project_group_assignment
ChatGPTManage group assignments for projects. Use get_all_projects to retrieve project IDs and get_all_groups to retrieve group IDs. Can add groups, remove groups, or replace all groups for a project. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
manage_project_location_assignment
ChatGPTManage location assignment for a project. Use get_all_projects to retrieve project IDs and get_all_locations to retrieve location IDs. Can assign or clear location for a project. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
manage_projects
ChatGPTManage projects by updating, archiving, or unarchiving them in batch. Use get_all_projects to retrieve project IDs. To clear optional fields: use '__CLEAR__' for string fields. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
manage_time_entries
ChatGPTEdits or deletes (archives) an existing clock-in, clock-out, or break time entry — use this to fix wrong or forgotten punches (e.g., 'I forgot to clock out yesterday'). Action 'update' changes the entry's time, note, activity, project, location, or coordinates. Action 'delete' removes the entry from the timesheet by archiving it. MANDATORY: First call 'get_time_entries' to find the exact time entry ID — NEVER guess or fabricate IDs. If multiple entries could match the user's request (e.g., several entries on the same day), ask the user which one to change before calling this tool. MANDATORY: For action 'delete', confirm with the user before calling this tool — deletion cannot be undone here. The entry's type (In/Out/StartBreak) and person CANNOT be changed: to fix a wrong entry type, delete the entry and recreate it via 'clock_in', 'clock_out', 'start_break', or 'end_break'. The backend rejects changes that break the punch sequence (e.g., two breaks in a row), duplicate an existing timestamp, fall into a locked pay period, or are not allowed by the workspace's edit permissions — report any backend error to the user. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
manage_time_off
ChatGPTManages a time-off request by updating its status (approve, reject, or cancel). To find the specific time-off request and its ID, use 'get_time_off_requests' first. Members can CANCEL their OWN requests (Pending or Approved) — the backend rejects attempts to approve or reject one's own request, which you must then report. Managers can approve/reject/cancel requests of people they manage; Admins and Owners can manage any request. Status rules enforced by the backend: Cancelled and Rejected requests cannot be changed; Approved requests can only be Cancelled. MANDATORY: Confirm with the user before cancelling or rejecting — these statuses are final and the request cannot be changed afterwards; a new request must be created via 'request_time_off' instead. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
manage_time_off_requests
ChatGPTApplies an explicit leave decision and returns the refreshed approval queue. For viewing, reviewing, filtering or paging requests, use get_time_off_requests_view instead. Calls without a decision remain supported for saved widgets. Shows the request breakdown by status (total, pending, approved, rejected) and a paginated table of requests with the person, dates, duration, policy, status and multi-level approval progress. Each pending row carries approve and reject controls the widget renders as buttons. Defaults to the current calendar month, which covers upcoming requests as well as past ones. Optional GroupIds or PersonIds narrow the queue; optional Status filters the table without changing the header counts. To decide on a request, pass LeaveId together with Decision (Approved | Rejected | Cancelled): the decision is applied and the refreshed queue is returned. Rejected and Cancelled are final and cannot be changed afterwards, so confirm with the user before sending either. Owners and admins only — for a member cancelling their own request, or a manager acting on one request conversationally, use 'manage_time_off'. This tool supports pagination. Use the 'PageSize' parameter to control the number of items returned per page (maximum: 100). Use the 'Page' parameter to request a specific page of results (1-indexed). The response includes: 'TotalCount' (total items), 'Page' (current page), 'PageSize' (items per page), 'TotalPages' (total pages), and 'ReturnedCount' (items in this response). To retrieve only counts without data, set 'PageSize' to 0. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first. When the widget is displayed, give only a brief summary without repeating its table or rows, unless the user explicitly requests a text table. If no widget is displayed, answer using the tool results.
request_time_off
ChatGPTCreates time-off request(s) for one or more people. For people, ONLY use Jibble person GUIDs/UUIDs; Member ID numbers are not valid person IDs. If the user supplied person names or Member IDs (not GUIDs), the AGENT MUST resolve person GUIDs via get_managed_people. If the user supplied a policy NAME (not an ID), the AGENT SHOULD resolve the GUID via get_all_time_off_policies; if multiple policies match, the AGENT SHOULD ask the user to choose. For full day or partial day leave, use StartDate (and optionally EndDate, inclusive). For hourly leave (Kind=Hours), use StartTime and EndTime with both date and time. When StartTime or EndTime is provided, values are submitted to the backend in UTC+0. When StartTime or EndTime is provided, the AGENT MUST call get_user_time_sheet_time_zone to determine the person's timesheet time zone before interpreting these values. For partial day leave (Kind=FullDay), provide PartOfDay and/or PartDayLeaveDuration. Maximum 6 items are allowed across array inputs in a single call. Do NOT bypass this by making multiple tool calls automatically. If the user provides more than this limit, ask the user to narrow the request first.
start_break
ChatGPTCreates StartBreak TE for person with given parameters. Supports starting a break at a specific past time using the time parameter. Future scheduling is not supported. To find most suitable location, use 'get_all_locations' tool. MANDATORY: Always call this tool when a user requests to start a break, regardless of previous conversation history or perceived status. This tool is the ONLY source of truth for the user's current break state. If the user is already on break, the backend will return an error (if any), which you must then report. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
update_organization_settings
ChatGPTUpdates specific settings for the organization. Settings to update as key-value pairs. Use get_settings_info tool to retrieve full settings list with settings names and value types. Duration settings (TimeSpan, e.g. ITimeTrackingSettings/AutomaticOutDelay, IScheduleSettings/ClockInBeforeStart) take ISO 8601 durations such as 'PT8H' or 'PT1H30M'; 'hh:mm:ss' is accepted and converted. Use '__CLEAR__' to clear a setting. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
update_schedule
ChatGPTUpdates an existing schedule. Only provided fields will be updated; omitted fields retain existing values. Use get_all_schedules to resolve schedule ID if schedule name is provided instead. This tool does not support batching. Do NOT call this tool multiple times automatically to fulfill one user request. If the user asks for repeated actions with this tool, ask them to split and send separate requests.
add_activities
Claudeadd_clients
Claudeadd_hour_entries
Claudeadd_projects
Claudeadd_schedule
Claudeassign_person_to_position
Claudeassign_schedules
Claudeclock_in
Claudeclock_out
Claudedelete_schedules
Claudeend_break
Claudeget_all_activities
Claudeget_all_clients
Claudeget_all_groups
Claudeget_all_locations
Claudeget_all_positions
Claudeget_all_projects
Claudeget_all_schedules
Claudeget_all_time_off_policies
Claudeget_api_docs
Claudeget_app_ui_navigations
Claudeget_article
Claudeget_attendance_insights
Claudeget_clock_in_patterns
Claudeget_geofence_compliance_dashboard
Claudeget_group_persons
Claudeget_group_settings
Claudeget_leave_balance
Claudeget_leave_history
Claudeget_leave_utilization
Claudeget_managed_people
Claudeget_my_work_summary
Claudeget_organization_settings
Claudeget_overtime_analysis
Claudeget_people_by_schedule
Claudeget_person_assignments
Claudeget_person_work_summary
Claudeget_public_holidays
Claudeget_report_export
Claudeget_settings_info
Claudeget_subscription_feature_catalogue
Claudeget_team_attendance_dashboard
Claudeget_team_schedule_view
Claudeget_team_time_insights
Claudeget_time_entries
Claudeget_time_off_requests
Claudeget_time_off_requests_for_groups
Claudeget_time_tracking_report
Claudeget_timesheets
Claudeget_timesheets_for_groups
Claudeget_user_settings
Claudeget_user_time_sheet_time_zone
Claudeget_who_clocked_in
Claudeget_who_clocked_out
Claudeget_who_on_break
Claudeget_wizard
Claudeget_wizards
Claudemanage_activities
Claudemanage_activity_group_assignment
Claudemanage_groups
Claudemanage_people
Claudemanage_positions
Claudemanage_project_client_assignment
Claudemanage_project_group_assignment
Claudemanage_project_location_assignment
Claudemanage_projects
Claudemanage_time_entries
Claudemanage_time_off
Clauderequest_time_off
Claudestart_break
Claudestart_report_export
Claudeupdate_organization_settings
Claudeupdate_schedule
Claude
