Menu

Individual skill

operating-infra

Author, inspect, troubleshoot, review, and apply (after explicit confirmation) infrastructure across IaC, Kubernetes, cloud resources, containers, CI/CD, and Linux hosts. Use when changing Terraform/OpenTofu, Kubernetes, Helm, Kustomize, Dockerfiles, GitHub Actions workflow/job/permissions semantics, AWS, GCP, Cloud Run, BigQuery, IAM, logs, instances, or service health, or when the user says "deploy", "deploy to staging", "terraform apply", "helm upgrade", "kubectl apply", "rollout", "deploy check", "validate deployment", or "validate infrastructure". NOT for shell scripts, generic command pipelines, or only the shell body inside `run:` steps (see writing-shell).

src/skills/operating-infra/SKILL.md

Read pinned source

Source commit 5ad5b8f0df027cf694ccbeb3311830890f01d582

What was checked

Matches the observed repository HEAD
Yes
Scanner rules are current
Yes
Repository scan complete
No
Skill scan complete
No
Observed behavior
Local scripts
Last checked
Nov 7, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings)
38
Repository last updated
Oct 25, 2026
License
MIT

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files
Read excerpt

# Operate Infrastructure

Work from files, plans, logs, and read-only commands; edit repo files freely, but touch live resources only under these rules:

- Before any cloud command, confirm identity (`aws sts get-caller-identity --profile <profile>`; `gcloud auth list`, `gcloud config list`), passing profile, project, region, and zone explicitly instead of relying on CLI defaults.
- Before any live change that's destructive, costly, or externally visible (apply, upgrade, rollout, delete, destroy, stop, resize, scale, IAM, bucket, network, DDL/DML, rollback): show identity, exact resources (ARNs or names), blast radius, irreversibility, and the plan/diff/inventory behind them, then wait for explicit confirmation.
- Every apply, upgrade, or rollout, regardless of blast radius: confirm the exact destination first (account, context, namespace, workspace, or release — name production explicitly), run the validation gates below on the same rendered artifact, show the plan or diff with create/modify/delete counts, and apply only that same reviewed artifact (the saved plan file or rendered manifest), only after explicit confirmation of that exact artifact and destination — never apply to production without it.
- After applying, verify rollout status, pod health, or Terraform outputs/state, and name the rollback path. On apply failure or a timed-out/degraded rollout: stop, report status and rollback options, and ask before any rollback.
- Without write access, return proposed changes (file, change, reason) instead of applying them.

For troubleshooting: rank likely causes, gather one safe signal at a time, propose the next step. For authoring: pick the smallest pattern keeping ownership, state boundaries, and least privilege.

In GitHub Actions this skill owns workflow structure, triggers, permissions, runners, actions, environments, secrets, caching, and concurrency — not the shell body of a `run:` step (writing-shell); mixed changes use both.