Appearance
Complete API Reference
All endpoints are hosted under /api/v1 and return standard JSON responses.
Ingestion Endpoints
1. Ingest Span Batch
- Method:
POST - Path:
/api/v1/traces/ingest - Auth: API Key (
X-API-Key) or Bearer Token - Request Body:
json
{
"resourceSpans": [
{
"resource": {
"attributes": [{ "key": "service.name", "value": { "stringValue": "payment-service" } }]
},
"scopeSpans": [
{
"spans": [
{
"traceId": "4bf92f3577b34da6a3ce929d0e0e4736",
"spanId": "00f067aa0ba902b7",
"parentSpanId": "5fb397be34d23b0f",
"name": "POST /charge",
"kind": 1,
"startTimeUnixNano": "1719200000000000000",
"endTimeUnixNano": "1719200000250000000",
"status": { "code": 1 }
}
]
}
]
}
]
}- Response:
202 Acceptedwith{ "ingested_spans": 1, "status": "queued" } - Error Codes:
400 Bad Request(invalid schema),429 Too Many Requests(buffer full).
Query Endpoints
2. Fetch Trace Detail & Spans
- Method:
GET - Path:
/api/v1/traces/{trace_id} - Response:
200 OKwith full list of spans, metadata, and root span summary.
3. Compute Critical Path
- Method:
GET - Path:
/api/v1/traces/{trace_id}/critical-path - Response:
200 OKwith ordered list of critical path span IDs and total critical duration.
4. Service Dependency Graph
- Method:
GET - Path:
/api/v1/dependencies - Response:
200 OKwith adjacency matrix: caller, callee, call count, and p95 latency.
Machine Learning & AI Endpoints
5. Anomaly Detection
- Method:
GET - Path:
/api/v1/ml/anomalies?window_minutes=60 - Response:
200 OKwith list of flagged services, baseline latency, observed latency, and z-score.
6. Automated Incident Summary
- Method:
POST - Path:
/api/v1/ml/summarize-incident - Request Body:
{ "trace_id": "4bf92f3577b34da6a3ce929d0e0e4736" } - Response:
200 OKwith executive summary, root cause analysis, and actionable remediation steps.