instatus_acknowledge_incident
ChatGPTAcknowledge an incident page as the signed-in user, the same as tapping Acknowledge in the app. It cannot acknowledge for someone else. The first acknowledgement cancels any escalation still waiting to page more people and makes the user the incident lead if there is none. Nothing changes on the public status page and no subscribers are notified. Only call this when the user says they are taking the incident. Returns who has acknowledged and who leads the incident.
instatus_add_incident_update
ChatGPTPublish a new message on an existing incident. On a published incident the message appears on the public status page immediately and, when notify_subscribers is true, subscriber notifications are queued; on an unpublished one it stays in the dashboard until the incident is published. It can also change the incident status and component statuses. To correct a message already posted use instatus_edit_incident_update; to close an incident use instatus_resolve_incident. Returns the new update id plus publication and notification status.
instatus_add_maintenance_update
ChatGPTPublish a progress message on an existing maintenance and optionally move its status (INPROGRESS starts it and shows its components as under maintenance, COMPLETED ends it). Completing sets each component back to OPERATIONAL unless an open incident or another running maintenance still covers it; those go back to the status that incident or maintenance asks for, listed in components_kept_affected. The message is public immediately; notifications are queued only when notify_subscribers is true.
instatus_add_subscriber
ChatGPTSubscribe an email address, phone number or webhook URL to a status page. With auto_confirm true the subscription is active immediately and no confirmation message is sent, on paid plans; on other plans auto_confirm_ignored comes back true and Instatus has sent a confirmation the recipient must accept. The confirmation field reports what actually happened. SMS needs SMS subscriptions turned on for the page. A webhook also needs webhook_email, the address told when deliveries fail, which is also where its confirmation goes. confirmation_sent_to names who received it. Only add contacts the user explicitly provided.
instatus_cancel_maintenance
ChatGPTCancel a maintenance by permanently deleting it. It disappears from the public page and history, this cannot be undone, and subscribers are not told. To tell them, post with instatus_add_maintenance_update first. Each component goes back to what any open incident or other running maintenance asks for, otherwise to operational; components reports the result.
instatus_create_component
ChatGPTAdd a component (a service customers see) to a status page, starting as operational. It appears on the public page immediately. Pass group_id to put it inside a group. A group is simply a component that has other components inside it: create it, then add or move components into it.
instatus_create_incident
ChatGPTOpen a new incident on a status page with its first message. By default this PUBLISHES the incident on the public page immediately; set publish to false to create it unpublished (visible only in the dashboard, and then notify_subscribers must be false). Queues subscriber notifications only when notify_subscribers is true. Write public text in the language of the page (language from instatus_list_pages), or ask the user. To add a message to an incident that already exists use instatus_add_incident_update. To log an incident that is already over, set status to RESOLVED with started_at and resolved_at: the history shows its real length and severity while live component statuses stay as they are. Returns the incident id, URL, publication and notification status, plus similar_open_incidents when one already covers these components. An open incident with the same title on the same components is refused as a likely double send unless allow_duplicate is true.
instatus_delete_component
ChatGPTPermanently remove a component from a status page, along with its uptime history. This cannot be undone. Confirm with the user first, unless the component was only just created in this conversation.
instatus_delete_incident
ChatGPTPermanently delete an incident and all its updates, for one opened by mistake. This cannot be undone, it removes the incident from the public page and history, and subscribers are not told. Each component it affected goes back to what any other open incident or maintenance asks for, otherwise to operational; components reports the result. To close a real incident use instatus_resolve_incident instead.
instatus_edit_incident_update
ChatGPTReplace the text of an update already posted on an incident. The new text is public immediately, keeps its original time and status, adds no new update and never notifies subscribers. Get the update id from instatus_get_incident. To add a new message instead use instatus_add_incident_update.
instatus_get_incident
ChatGPTRead one incident in full: title, current status, affected components, public URL and every update in order (newest first). Read-only. Use this before writing an update so the new message builds on what was last published. paging says whether the on-call team was paged and who acknowledged; its lead is whoever the incident is assigned to, which can be set without anyone being paged. Acknowledge with instatus_acknowledge_incident.
instatus_get_maintenance
ChatGPTRead one maintenance window with all of its updates. Read-only.
instatus_get_monitor_checks
ChatGPTRead a monitor’s latest check results, newest first, to explain why it is down or flapping: HTTP status, response time, the location that ran each check, retries and any failed assertions (the rule and the value it got). Read-only. diagnosis summarises the checks read (with failures_only, only the failure counts): when the latest check ran (compare it with the check interval: an old latest check means the data is stale), whether it passed, failing_since (the oldest failure in the current run of failures, or the oldest check read when no success was read), last_success_at, and failures by HTTP status and by location. Set failures_only to read only failed checks, and since / until to look at a past episode instead of the latest checks.
instatus_get_page_status
ChatGPTAnswer “is anything broken on this status page right now?” in one call. Read-only. Returns the page’s overall status worked out from its live components, the components that are not operational, the names of the ones that are, active (unresolved) incidents with their latest update, and scheduled or in-progress maintenances, plus observed_at. A component in neither list does not exist on this page. overall_status names the worst level only: ONEPARTIALOUTAGE means one component at that level, and others may be affected less badly, so use summary for the count. An unpublished incident is visible only in the dashboard and its URL does not work publicly. Sections that failed to load are listed in missing_sections; an empty list there means the data is complete. For full incident history use instatus_get_incident.
instatus_get_uptime
ChatGPTMeasure uptime and response times for monitors over a period: uptime_percent (share of successful checks), estimated_downtime_minutes (failed checks times the check interval, as the dashboard counts it), failed checks, average response time, and the 95th percentile (p95_response_ms for periods of five days or less; past that only avg_daily_p95_response_ms, the average of daily p95s, which is not the period’s p95). Read-only. Give monitor_id for one monitor, or leave it out for every monitor on the page (up to 25, lowest uptime first). Paused monitors are measured too, since they can have checks for a past period. The period defaults to the last 30 days and can be up to 395 days; pass start and end as ISO 8601 times. Periods longer than five days are measured in whole days, and an end at midnight stops the day before, so 1 October to 1 November 00:00 is exactly October; start and end in the result are the period actually measured, and period_note says when that differs from the one asked for. When a monitor has no checks for the start of the period, data_starts and coverage_note say so: never present such figures as covering the whole period. Set daily for a per-day breakdown. uptime_percent is null when a monitor ran no checks in the period. This measures the monitors, not the components shown on the status page.
instatus_list_components
ChatGPTList every component (service) shown on a status page with its exact id, current status and group. Read-only. is_group marks a group that holds other components, whose status is that of its worst child, and group_id links a component to its group. Archived components are excluded unless include_archived is true. Use these ids in affected_components of incident tools.
instatus_list_general_notices
ChatGPTList the general notices on a status page: banners shown at the top of the page that are not an incident or a maintenance (for example a pricing change or a holiday support schedule). Read-only. Every notice listed is currently shown on the page.
instatus_list_incidents
ChatGPTList incidents on one status page, newest first. Read-only. By default returns only unresolved incidents; set include_resolved to see history, and since to limit it to a period such as this month. When next_cursor is not null there are more: pass it back as cursor. Each item carries the exact incident id, title, status, public URL, start and resolve times, duration_minutes once resolved, ongoing_minutes while open, impact and the latest update. For the full update timeline of one incident use instatus_get_incident.
instatus_list_maintenances
ChatGPTList scheduled maintenance windows on a status page. Read-only. By default returns upcoming and in-progress maintenances; set include_completed for history. Each item has the exact maintenance id, title, status, planned starts_at and ends_at, actually_started_at and actually_ended_at, and affected components. duration_minutes is the planned length until it finishes, then the real one. When next_cursor is not null there are more: pass it back as cursor.
instatus_list_monitors
ChatGPTList the uptime monitors on a status page (website, API, ping, TCP, UDP and DNS checks) with each one’s current status: UP, DOWN, DEGRADED or UNKNOWN. Read-only. counts says how many are down, degraded or paused. Paused monitors keep their last status but are not checking. For why a monitor is down, use instatus_get_monitor_checks; for uptime over time, instatus_get_uptime. Cron (heartbeat) monitors are not included; they are in the Instatus dashboard.
instatus_list_outages
ChatGPTList past and ongoing periods when components on a status page were not operational, ongoing ones first, then newest first: which component, the status it had (for example MAJOROUTAGE or UNDERMAINTENANCE), when it started and ended, how many minutes it lasted, and the incident behind it if any. Read-only. Use it for questions like "what went down last week" or "how long was the API out". Pass since to stop at a date and component_id for one component.
instatus_list_pages
ChatGPTFind the status pages this connection can access, sorted by name. Read-only. Returns each page’s exact id (use it in every other tool), name, public URL, subdomain, workspace and creation date so similarly named pages can be told apart. Each page’s status can trail recent changes; use instatus_get_page_status for the current state. connection_access says whether this connection can change anything; page_access is selected_pages when the user approved it for some pages only, so a page they mention may be missing here because it was not approved, not because it does not exist. Pass query to filter by name, URL or workspace. If more than one page matches the user’s description, ask the user which one instead of picking the first; if next_cursor is not null, the list is incomplete.
instatus_list_subscribers
ChatGPTList people and endpoints subscribed to a status page. Read-only. Returns personal contact data; only call it when the user explicitly asks about subscribers. For “how many subscribers” set count_only, which returns totals without any contact data. Supports search by email or phone. When next_cursor is not null there are more: pass it back as cursor. A search can return no matches with a next_cursor because it has more subscribers left to check; keep going until next_cursor is null before saying a contact is not subscribed.
instatus_post_general_notice
ChatGPTShow a general notice on a status page: a banner that is not tied to an incident or a maintenance. It is PUBLIC on the page immediately and stays until it is removed. Use instatus_create_incident for something broken and instatus_schedule_maintenance for planned work. Queues subscriber notifications only when notify_subscribers is true. Write public text in the language of the page.
instatus_remove_general_notice
ChatGPTTake a general notice off the status page permanently. This cannot be undone. Only call this when the user asked to remove that notice.
instatus_remove_subscriber
ChatGPTPermanently remove one subscriber from a status page. They stop receiving updates at once and are not told. This cannot be undone. Get the id from instatus_list_subscribers with search, and only call this when the user explicitly asked to remove that contact.
instatus_resolve_incident
ChatGPTClose an incident by publishing a final RESOLVED update. The message appears publicly right away; notifications are queued only when notify_subscribers is true. By default every component the incident affected is set back to OPERATIONAL, except components another open incident still affects, which keep their current status (set restore_components to false to leave every status as it is). Returns the resolved incident, which components were restored and which were kept affected.
instatus_schedule_maintenance
ChatGPTCreate a scheduled maintenance window on the public status page. It is visible on the page immediately as upcoming. Instatus starts and ends it automatically at the given times when auto_start / auto_end are true. A window already in the past is recorded straight away as completed; check with the user that this is what they meant. Notifications: notify_subscribers controls the announcement now; notify_on_start and notify_on_end control notifications at those moments. Returns the maintenance id and schedule.
instatus_set_component_status
ChatGPTChange the displayed status of one or more components directly, without creating an incident or posting any message. The new status is public immediately and subscribers are not notified. For customer-facing communication open an incident with instatus_create_incident, or for planned work schedule one with instatus_schedule_maintenance. Check open incidents first: this overwrites what they show. Components change one at a time, so a failure part way leaves the earlier ones changed: components lists those, failed says whether that one changed, and not_attempted lists the rest. Returns previous_status for each, and still_covered_by lists any open incident or running maintenance that says otherwise, which stay public until they are resolved or completed.
instatus_update_component
ChatGPTRename a component, change its description, or move it in or out of a group. The change is public immediately. To change its status use instatus_set_component_status.
instatus_update_general_notice
ChatGPTChange the title, message or collapsed state of a general notice. The change is public immediately. No notifications are sent. Only the fields you pass change.
instatus_update_incident
ChatGPTEdit an existing incident’s title or start time, or publish an unpublished incident (publish: true, which applies its component statuses to the public page). This does not post a message or change status. It never notifies subscribers, unless publish and notify_subscribers are both true. To post a message use instatus_add_incident_update; to correct one use instatus_edit_incident_update. Returns the saved values.
instatus_update_maintenance
ChatGPTReschedule or reword a maintenance that has not finished: change its title, times, affected components or the text of its first message. The change is public immediately and subscribers are not notified. Times keep the same length unless both are sent. To post a new progress message use instatus_add_maintenance_update.
instatus_who_is_on_call
ChatGPTShow who is on call right now and who is next, for each on-call schedule of a status page. Read-only. A schedule with nobody in on_call_now has a gap in its rotation, or is paused. Returns names and work emails of team members.