How osnabrugge/home-ops works
This repository is a GitOps-managed home-infrastructure estate: operators maintain Talos and Kubernetes configuration, bootstrap the cluster’s critical dependencies, and Flux reconciles the remaining applications from the repository. The main result is a running homelab with networking, storage, access, backup, and monitoring capabilities. The graph groups runtime infrastructure by responsibility and includes the documented AI-triage pilot. Relationships are limited to README-described flows; application internals and most wiring were not sampled, so those connections are omitted.
Main entry points: apps.yaml Bootstrap Chain · flux-system Flux Reconciliation · litellm AI Model Routing · kopiur PVC Backups
Components 26
Cluster Bootstrap
- Bootstrap Chain
bootstrap/helmfile/apps.yaml - Talos Nodes
talos/nodes - Flux Reconciliation
kubernetes/apps/flux-system
Cluster Platform
- Cluster Networking
kubernetes/apps/network - Cluster DNS
kubernetes/apps/kube-system - Certificates
kubernetes/apps/cert-manager - Secret Management
kubernetes/apps/external-secrets - Cluster Storage
kubernetes/apps/rook-ceph - Local Storage
kubernetes/apps/openebs-system - External Access
kubernetes/apps/network/cloudflare-tunnel
Operations
- PVC Backups
kubernetes/apps/kopiur-system/kopiur - Garage S3
- Monitoring Stack
kubernetes/apps/observability/kube-prometheus-stack - Alert Triage
kubernetes/apps/observability/alert-triage/app/triage.py
User Workloads
- Home Applications
kubernetes/apps/default - Identity Services
kubernetes/apps/auth
AI Workflows
- AI Model Routing
kubernetes/apps/ai/litellm - Local Models
kubernetes/apps/ai/ollama - Read-only Triage
scripts/triage_readonly.py - Scheduled Triage
kubernetes/apps/ai/triage-worker/app/worker.py - Kritika Pilot
kubernetes/apps/kritika-system/kritika/app/policy.py
Other components
- Home Operator
- Infrastructure Repo
- GitHub
- Azure Key Vault
- Kopia / Garage S3
Connections 21
- Home Operator → Infrastructure Repo: maintains
README.md - Infrastructure Repo → Talos Nodes: configures
README.md - Infrastructure Repo → Bootstrap Chain: defines
README.md - Bootstrap Chain → Cluster Networking: installs first
README.md - Bootstrap Chain → Cluster DNS: installs next
README.md - Bootstrap Chain → Certificates: installs after
README.md - Bootstrap Chain → Secret Management: installs after
README.md - Bootstrap Chain → Flux Reconciliation: installs after
README.md - Infrastructure Repo → Flux Reconciliation: reconciled by
README.md - Flux Reconciliation → Home Applications: reconciles
README.md - Flux Reconciliation → Cluster Storage: reconciles
README.md - Flux Reconciliation → Monitoring Stack: reconciles
README.md - Flux Reconciliation → PVC Backups: reconciles
README.md - PVC Backups → Kopia / Garage S3: backs up to
README.md - Monitoring Stack → Alert Triage: alerts feed
- Azure Key Vault → Secret Management: supplies secrets
README.md - External Access → Home Applications: exposes services
- Read-only Triage → AI Model Routing: uses model routing
README.md - Scheduled Triage → AI Model Routing: uses model routing
- AI Model Routing → Local Models: routes to
- GitHub → Kritika Pilot: review pilot