name	harness
description	Route repeated MoruPark infrastructure and Kubernetes operations through the local harness documentation and skills. Use when Codex needs to choose between cluster reprovisioning, ESO recurrence prevention, Kubernetes release verification, or broader infra operations, and when a request benefits from the standard runbook order and repo-specific guardrails before acting.

Harness

Use this meta skill to choose the right repo-local operations skill and follow the standard harness order.

Read references/skill-routing.md to choose the correct skill.
Read references/operating-order.md when the task touches prod infra or cluster recovery.
Dispatch to the underlying skill with the smallest matching scope.

Use eso-recurrence-prevention for ExternalSecret, SecretStore, CRD, or synced Secret failures.
Use cluster-reprovision-handoff after Terraform reprovisioning, cluster recreation, or kubeconfig re-bootstrap work.
Use k8s-release-verification for rollout health, overlay verification, and post-deploy checks.
Use infra-ops for broader Terraform-centric infra work that is not primarily ESO or rollout verification.

Do not jump into prod mutation before identifying the correct operational path.
Do not treat cluster rebuild, ESO bootstrap, and app rollout as the same task.
Do not skip the repo-standard order when the request involves reprovisioning.
Record important completed Codex/skill-driven ops work in infra/md/implement.md.

name	harness
description	Route repeated MoruPark infrastructure and Kubernetes operations through the local harness documentation and skills. Use when Codex needs to choose between cluster reprovisioning, ESO recurrence prevention, Kubernetes release verification, or broader infra operations, and when a request benefits from the standard runbook order and repo-specific guardrails before acting.

Use this meta skill to choose the right repo-local operations skill and follow the standard harness order.

Read references/skill-routing.md to choose the correct skill.
Read references/operating-order.md when the task touches prod infra or cluster recovery.
Dispatch to the underlying skill with the smallest matching scope.

Use eso-recurrence-prevention for ExternalSecret, SecretStore, CRD, or synced Secret failures.
Use cluster-reprovision-handoff after Terraform reprovisioning, cluster recreation, or kubeconfig re-bootstrap work.
Use k8s-release-verification for rollout health, overlay verification, and post-deploy checks.
Use infra-ops for broader Terraform-centric infra work that is not primarily ESO or rollout verification.

Do not jump into prod mutation before identifying the correct operational path.
Do not treat cluster rebuild, ESO bootstrap, and app rollout as the same task.
Do not skip the repo-standard order when the request involves reprovisioning.
Record important completed Codex/skill-driven ops work in infra/md/implement.md.

harness