Observability & Request Trace

Investigate a single request from gateway acceptance through identity, policy, approval, budget, execution, usage, and audit outcome.

Why request-centric observability matters

Traditional gateway metrics answer fleet questions such as latency, spend, or provider health. Incident response usually starts with a different question: what happened on this exact request?

Syndicate Gate correlates gateway telemetry around a stable request ID so operators can move from a support ticket, audit review, or policy investigation to the precise execution trace without reconstructing it from unrelated logs.

Hosted customer dashboard

Hosted deployments expose a customer-facing observability view at /dashboard/observability. Operators can search by request ID and review:

audit timeline and request status
routing path and selected provider
guardrail evaluations and policy outcomes
token usage and computed cost context

Enterprise admin dashboard

Enterprise deployments expose the same investigative model through the admin console request trace surface at /console/#trace. This view is intended for self-hosted operators who need request reconstruction from the local ledger and gateway event stream.

API surface

Hosted request detail is available through the admin API using the request-specific endpoint below:

GET /api/v1/admin/requests/:request_id

The response is designed for request investigation rather than bulk analytics. It returns summary, audit timeline, routing detail, guardrail evaluations, and cost fields when available.