Check per-service health

This endpoint is public, so send no token. It returns the same ok and traffic-light status as GET /gateway/health, and adds a services array for operations dashboards and for working out where a problem sits. Each entry carries the service's name, important (whether it counts as critical), and ok (whether it was reachable). The breakdown covers the internal services the gateway coordinates, which are authentication, vaults and keys, tokens and contracts, spending policy, audit logging, and notifications, so expect more entries than the sample below shows.

The traffic light follows the breakdown. Green means every service reported ok. Orange means a service with important: false is down while the core guarantees still hold. Red means a service with important: true is down and value-moving actions cannot be processed safely. Read the two together: the status tells you whether to hold movements, and the breakdown tells you which service to chase. Results are cached for about 15 seconds, so a dashboard refreshing faster than that will see the same payload.