Kube-Hetzner is an infrastructure project that uses Terraform or OpenTofu to deploy and maintain Kubernetes clusters on Hetzner Cloud. It is for operators who need highly available clusters based on k3s or RKE2, with networking, storage, ingress, certificates, and upgrades configured as part of the deployment.
Borrowing it
Nothing to install: this file belongs to mysticaltech/terraform-hcloud-kube-hetzner. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/mysticaltech/terraform-hcloud-kube-hetzner/master/.claude/skills/running-stabilization-loop/SKILL.mdgit clone --depth 1 https://github.com/mysticaltech/terraform-hcloud-kube-hetznerWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mysticaltech/terraform-hcloud-kube-hetzner/running-stabilization-loop)<a href="https://agentmods.dev/skills/mysticaltech/terraform-hcloud-kube-hetzner/running-stabilization-loop"><img src="https://agentmods.dev/badge/skills/mysticaltech/terraform-hcloud-kube-hetzner/running-stabilization-loop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mysticaltech/terraform-hcloud-kube-hetzner/running-stabilization-loop"><img src="https://agentmods.dev/badge/skills/mysticaltech/terraform-hcloud-kube-hetzner/running-stabilization-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00086 | $0.03788 |
| Opus 5 | $0.00043 | $0.01894 |
| Sonnet 5 | $0.00017 | $0.00758 |
| Haiku 4.5 | $0.00009 | $0.00379 |
Grade A, and why
running-stabilization-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 399 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Running Stabilization Loop
Overview
Use this workflow when the goal is not only to detect failures, but to fix them and prove full regression safety after major changes.
Usage
/running-stabilization-loop
Buddy Prompt Template
Use this exact prompt when delegating the loop to another agent:
Run the full kube-hetzner v3 stabilization loop end-to-end, do not stop at pass/fail reporting.
Goals:
1) Make every matrix scenario pass from simple to complex.
2) Fix root causes in kube-hetzner (not workarounds in test harness) unless explicitly required.
3) Re-run full matrix after fixes to prove no regressions.
4) Validate upgrade path from latest git tag to the current release-candidate branch for both k3s and rke2.
5) When Tailscale credentials are available, include the Tailscale node-transport
scenarios: secure single-network Tailnet access, plus Flannel with multiple
Hetzner Networks, route auto-approval, autoscaler scale-up/down,
`tailscale status`, and cross-network pod traffic.
6) Include final v3 polish scenarios when touched: Cilium Gateway API with
`enable_kube_proxy=false`, embedded registry mirror on k3s or RKE2, endpoint
output sanity, and example validation.
Hard requirements:
- Use /Users/karim/.ssh/id_ed25519 with IdentitiesOnly=yes for all SSH.
- Use /Volumes/MysticalTech/Code/kube-test/run_v3_matrix.sh and summary-*.md as source of truth.
- If RKE2/K3s apply hangs, SSH during apply and inspect journalctl/cloud-init immediately.
- If destroy hangs, run cleanup and delete stuck autoscaled servers via hcloud.
- Keep changes minimal and production-safe; run Terraform and OpenTofu validation before each rerun.
- For Tailscale node transport, verify single-network clusters work without
route approval when route advertisement is disabled; for multinetwork, verify
node-private routes are approved and SNAT is disabled for advertised routes.
Do not mistake external overlay SSH access for supported Kubernetes node
transport.
- For Cilium Gateway API, verify GatewayClass exists, Gateway is accepted, and
HTTPRoute serves through the Hetzner LoadBalancer. For embedded registry
mirror, verify `registries.yaml`, `embedded-registry: true`, peer ports, and
image-pull behavior.
Deliverables:
- Final scenario matrix results with log paths.
- List of fixes made with files changed.
- Upgrade test result (tag -> release-candidate branch) and any required migration notes.
- Confirmation cleanup is complete and no khv3-* resources are left.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago Changed · -4 lines d2d38b4b8347
- 10d ago First seen · 403 lines · 86 tokens per session scan E d9efc638206e
running-stabilization-loop is a skill published in the GitHub repository mysticaltech/terraform-hcloud-kube-hetzner (3,927 stars, last pushed 2d ago), licensed MIT. It adds 86 tokens to every session and 3,788 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
api-harness
Use the API harness to drive Agentweaver through its REST API, capture the complete request/response evidence, and emit a normalized agentweaver.persona-judge-verdict/v1 JSON verdict. It is for backend/API end-to-end validation; use the UI or MCP harness for those surfaces.
ui-harness-cli
Run the Agentweaver persona-driven Playwright UI evidence harness. Use for a specific persona's deployed browser flow, end-to-end UI validation, or investigation of a UI-reported issue.
persona-briefs
Use this contract when you need to discover the current built-in scenario/persona catalog for one harness surface, or generate a reviewed new persona core and surface adapter for a new test intent.
Kubernetes Testing Patterns
Kubernetes deployment testing including manifest validation, Helm chart testing, pod health checks, network policy testing, and rollback verification.
k8s-aiops
Use this skill whenever the user needs to operate a Kubernetes cluster — list/inspect pods, deployments, statefulsets, daemonsets, replicasets, jobs, cronjobs, services, ingresses, endpoints, configmaps, secrets (names/keys only), PVCs/PVs/storageclasses, nodes, namespaces, and events; read pod logs; describe…
dogfood
Systematically explore and test a mobile app on iOS/Android with agent-device to find bugs, UX issues, and other problems. Use when asked to dogfood, QA, exploratory test, find issues, bug hunt, or test this app on mobile.