Audit chain integrity
Checking ...
- Head hash
- not yet measured
- Total entries
- not yet measured
- Last verified at
- not yet measured
JWKS availability
Checking ...
- Active keys
- not yet measured
- Last verified at
- not yet measured
API health
Checking ...
- Reported status
- not yet measured
- Last verified at
- not yet measured
Components
| Component | Status | Note |
|---|---|---|
Pre-policy hook | Unknown | Synchronously evaluated at every service_proxy hot path. Reported via the API health probe; deeper per-rule latency is not yet exposed on the public surface. |
Joint-action evaluator | Unknown | Runs synchronously alongside the pre-policy hook and remains the default; this blocks the action at the boundary. Surfaces via the same API health signal. An opt-in async path exists behind a default-off flag and records a governance outcome after the action lands rather than blocking it; it is not yet the operational default. |
Audit chain | Unknown | Head pointer read from the public verify endpoint. Green means the chain has advanced and the latest entry hash is readable; offline verification against the JWKS remains the authoritative check. |
IAT signing | Unknown | Public JWKS endpoint is reachable and reports at least one active signing key. Token issuance latency is not yet measured on the public surface. |
Database | Unknown | PostgreSQL reachability as reported by the API health probe (checks.database). Connection-pool detail is on the internal platform-health dashboard, not the public surface. |
Redis | Unknown | Cache backend reported by the API health probe (checks.cache). Green covers a connected Redis or the in-memory fallback; the probe distinguishes the two. Kill-switch propagation depends on this layer and fails safe to enforce on a cold cache. |
Workers | Unknown | Background workers (Celery) are not yet exposed on the public health probe, so their status is reported as unknown here rather than assumed healthy. Worker health is observed internally; a public signal is on the roadmap. |
Adapter sandboxExperimental | Degraded | CPU, memory, and wall-clock resource limits are enforced, along with filesystem and network confinement where the host runtime supports it; a fallback path without that support enforces neither. See the honest limits page. |
Recent incidents
No incidents recorded since launch.
This is a placeholder for the incident log format. A durable incident-recording mechanism is on the roadmap; until it ships, this section is intentionally honest about being empty rather than backfilling fiction.
Signals on this page are pulled from public Vortalis endpoints. The audit chain is independently verifiable offline against the JWKS; this page is a convenience surface, not a replacement for that verification. SLOs and uptime percentages are deliberately omitted because they are not yet measured on the public surface.
Read what Vortalis does not defend against