Monitoring & Health
Squad exposes health and readiness endpoints for monitoring platform status, along with structured logging and correlation tracing for operational visibility.
Health & Readiness
The platform provides standard health check endpoints that verify connectivity to all dependent services:
| Endpoint | What It Checks |
|---|---|
GET /health | Basic platform liveness |
GET /health/ready | Full readiness: graph database, LLM service, cache, event store |
The readiness check verifies that all backing services are reachable and responsive. This is suitable for use as a Kubernetes readiness probe or load balancer health check.
Structured Logging
All platform components emit structured logs with consistent metadata:
- Correlation ID: Unique per-request identifier that flows through all service calls
- Session ID: Links log entries to a specific user session
- User identity: Authenticated user associated with the action
- Event type: Categorised event (query, retrieval, execution, review, error)
- Timestamps: ISO 8601 with millisecond precision
Correlation IDs are included in SSE event streams, allowing client-side tools to join real-time events with server-side logs.
Real-Time Event Stream
AIM’s processing pipeline emits Server-Sent Events (SSE) at each stage of query processing. This provides live visibility into what the system is doing:
- Classification results
- Disambiguation exchanges
- Retrieval operations
- Plan construction
- Step-by-step execution progress
- Review outcomes
Related
- Transparency & Audit: the immutable audit trail for compliance
- Analytics & Reporting: usage and performance analytics
- API Reference: full API documentation including health endpoints