Some systems are experiencing issues.
Source control, CI/CD pipelines, and the container-image build platform.
Container image registry with vulnerability scanning and signing.
Extension marketplace proxy for the GitLab Web IDE (Microsoft Marketplace).
Unified LLM/AI API gateway routing to the platform's models.
This public status page for the Neo platform.
Dashboards and log/metric/trace exploration (Grafana LGTM).
Kubernetes cost monitoring and FinOps reporting (EUR).
The metrics, logs, and traces storage behind the dashboards.
Team password manager and secrets vault (Bitwarden-compatible).
The runtime secrets service that issues application credentials.
Automated TLS certificate issuance and renewal for platform services.
Platform object storage for artifacts, backups, and application data.
Managed relational databases backing the platform's applications.
Scheduled backups of cluster workloads and their persistent data.
The infrastructure state store, with a warm standby and off-site copies.
Scheduled snapshots of the platform's secrets store.
Automated recovery drills that verify backups can be restored.
The platform's internet firewall and outbound connectivity.
The internal relay fabric that links the platform's clusters.
Internal name resolution for platform services.
Response time of Vaultwarden measured against the internal Neo-DNS service endpoint (the real backend, not the public edge).
Response time of Cachet measured against the internal Neo-DNS service endpoint (the real backend, not the public edge).
Response time of GitLab measured against the internal Neo-DNS service endpoint (the real backend, not the public edge).
Response time of Grafana measured against the internal Neo-DNS service endpoint (the real backend, not the public edge).
Response time of Harbor measured against the internal Neo-DNS service endpoint (the real backend, not the public edge).
Response time of LiteLLM measured against the internal Neo-DNS service endpoint (the real backend, not the public edge).
Response time of OpenCost measured against the internal Neo-DNS service endpoint (the real backend, not the public edge).
Response time of Harbor, measured from the status-page prober.
Response time of OpenCost, measured from the status-page prober.
Response time of Vaultwarden, measured from the status-page prober.
Response time of Grafana, measured from the status-page prober.
Response time of LiteLLM, measured from the status-page prober.
Response time of GitLab, measured from the status-page prober.
Response time of Cachet, measured from the status-page prober.
Response time of Web IDE Marketplace, measured from the status-page prober.
Enforced guardrail gates (self-props provers + doc/config gates), mined from the committed git history — the self-evolving system's growing safety envelope. Higher is more self-enforced.
Hard-won operational gotchas captured in the knowledge graph, mined from git history — accumulated, cross-linked institutional memory. Higher is more captured.
Architecture Decision Records, mined from git history — the documented, reviewable decision trail behind every structural change. Higher is more decisions recorded.
Deterministic code/doc generators (gen-*), mined from git history — the share of the platform that is generated-and-drift-gated rather than hand-maintained. Higher is more automated-by-construction.
Broken/dead internal doc references, mined from git history — the anti-bloat signal. This is the ONE series that should FALL: 12 → 0 shows the optimiser converging toward correctness, not just growing.
