Uptime: Wiki | Incident

Parameters

Path Parameters

id*integer

Entity id

Formatint64

Range0 <= value

Code example

curl -X GET "https://example.com/api/v1/incidents/0"

Response example

{  "id": 0,  "root_kind": "string",  "root_entity_id": 0,  "rack_id": 0,  "root_msg_key": "string",  "root_msg_args": [    {      "key": "string",      "text": "string",      "number": 0.1,      "is_number": true    }  ],  "symptom_kind": "string",  "symptom_msg_key": "string",  "symptom_msg_args": [    {      "key": "string",      "text": "string",      "number": 0.1,      "is_number": true    }  ],  "symptom_customer_id": 0,  "confirmed": true,  "causes": [    {      "root_kind": "string",      "entity_id": 0,      "rack_id": 0,      "msg_key": "string",      "msg_args": [        {          "key": "string",          "text": "string",          "number": 0.1,          "is_number": true        }      ],      "confidence": 0,      "confirmed": true,      "cleared": true,      "articles": [        {          "id": "string",          "title": "string",          "title_plain": "string"        }      ]    }  ],  "started_at_ns": 0,  "last_changed_at_ns": 0,  "resolved_at_ns": 0,  "affected_customer_ids": [    0  ],  "affected_host_ids": [    0  ],  "affected_host_total": 0,  "revenue_at_risk_per_hour_dollars": 0.1,  "max_sla_tier": "string",  "compliance_at_risk": "string",  "redundancy_promise_broken": true,  "band": "string",  "timeline": [    {      "at_ns": 0,      "kind": "string",      "detail": "string"    }  ],  "work_order_ids": [    0  ],  "channel_thread_id": 0,  "articles": [    {      "id": "string",      "title": "string",      "title_plain": "string"    }  ],  "hint_tier": 0,  "hint": "string",  "evidence": [    {      "label": "string",      "label_plain": "string",      "value": "string",      "value_plain": "string",      "detail": "string",      "detail_plain": "string",      "leading": true    }  ]}

Renderer-friendly mirror of sim_core::incident::Incident. Channel UI, post-mortem, and the "is this customer impacted?" badge all read from these. Severity (EventView.severity) is a per-event tag; band here is the incident-level gravity read.

id*integer

Formatint64

Range0 <= value

root_kind*string

Variant tag mirroring IncidentRoot — "HostFailed", "CableFailed", etc.

root_entity_id*integer

Primary entity id from the root, when there is one (host id, cable id, rack id, ...). 0 for global roots like Ddos.

Formatint32

Range0 <= value

rack_id?integer

Rack the incident beacon should mark — resolved engine-side from the typed root (device's rack, or the rack itself, or the edge rack where WAN ingress lands for global/fabric roots). 0 when nothing is racked yet. Lets the renderer skip a per-tick scan of every device array.

Formatint32

Range0 <= value

root_msg_key?string

Localization key for the root label — incident-<root_kind-kebab>.

Guidance ("what to check") deliberately has NO key field: it is static per root and the client derives guidance-<root_kind>-check-<n> from root_kind, which is already on this struct.

root_msg_args?array<>

Facts for root_msg_key, including any already-resolved customer name.

symptom_kind?string

Symptom layer — the customer/operator-facing HEADLINE the card leads with (many causes → one symptom). Variant tag mirroring Symptom: "ServiceUnreachable", "InfrastructureFault", etc.

symptom_msg_key?string

Localization key for the symptom headline — symptom-<kind-kebab>.

symptom_msg_args?array<>

Facts for symptom_msg_key.

symptom_customer_id?integer

Tenant the symptom is rooted on, or 0 for an InfrastructureFault.

Formatint32

Range0 <= value

confirmed?boolean

True if any candidate cause has been confirmed as the real cause.

causes?array<>

Ranked candidate causes, best-first. In Phase 2 exactly one confirmed candidate; full ranking is Phase 4.

started_at_ns*integer

Formatint64

Range0 <= value

last_changed_at_ns*integer

Formatint64

Range0 <= value

resolved_at_ns*integer

0 while open; set to the resolution sim-time once closed.

Formatint64

Range0 <= value

affected_customer_ids*array<>

Sorted ascending by id. Engine guarantees deterministic order.

affected_host_ids?array<>

The MACHINES this incident is about — sorted ascending, capped at sim_core::incident::AFFECTED_HOSTS_CAP. Empty for roots with no host scope (DDoS, uplink saturation, customer-observation): those degrade everyone, so naming hosts would be a lie. Resolve names against hosts[].display_name.

affected_host_total?integer

True count before the cap, so the UI can say "showing 64 of 812" rather than silently truncating.

Formatint32

Range0 <= value

revenue_at_risk_per_hour_dollars*number

Sum of monthly recurring run-rate across the affected set.

Formatdouble

max_sla_tier*string

Worst SLA tier in the affected set — "Bronze" / "Silver" / "Gold" / "Platinum".

compliance_at_risk*string

Worst compliance tier in the affected set — "—" / "SOC2" / "HIPAA" / "PCI".

redundancy_promise_broken*boolean

band*string

Computed gravity band — "cosmetic" / "annoying" / "serious" / "existential".

timeline?array<>

Incident timeline, oldest first. Bounded only by incident lifetime — preserved into the resolved-ring.

work_order_ids?array<>

Active work order ids targeting this incident. Populated once the work model lands (Pass A2); empty until then.

channel_thread_id?integer

Channel thread root in #incidents. Populated once the channel module lands (Pass A3); 0 until then.

Formatint64

Range0 <= value

articles?array<>

Knowledge-base articles covering this root's mechanic — the detail pane renders them as codex deep links. May be empty.

hint_tier?integer

Highest progressive-hint tier fired for this incident (0 = none, 1 = subsystem nudge, 2 = named suspect, 3 = named fix).

Formatint32

Range0 <= value

hint?string

The current-tier escalating hint text (empty when hint_tier == 0).

evidence?array<>

MEASURED evidence behind a customer-facing symptom: the tenant's p99 against its own baseline and tier budget, its failure rate, where the latency actually went, and why requests failed.

The card used to carry labels, prose and hints but not one number — its own guidance said "check their p99 latency vs tier budget" about a figure it never displayed. These rows are that figure. Empty for infrastructure-rooted incidents with no tenant to measure.

Incident

Parameters

Try the endpoint

Code example

Response example

Response body

On this page