メニュー
DR/HA discipline — RTO/RPO as requirements, DR tiers, backup verification, failover design. Cloud-agnostic.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Context engineering — memory navigation, textual graph traversal, progressive disclosure, recovery. Load when assembling context for a task or when lost.
Implementation-side security discipline — headers, dependencies, secrets, authz. Load for hardening work (sec-ops domain).
Commit conventions, branching, anchors in messages, and the git safety rules. Load for any git operation beyond status/diff.
Anti-generic-AI design discipline — bans, tells, and the slop test. Load for ANY new UI surface or visual redesign. Inspired by Impeccable (Apache 2.0).
Motion & animation discipline — purpose, performance, reduced-motion. Load when work involves animation, transitions or scroll effects.
Microcopy discipline — buttons, errors, empty/loading/success states, alt text. Voice and tone come from the project's memory, not from this skill.
| name | resilience |
| description | DR/HA discipline — RTO/RPO as requirements, DR tiers, backup verification, failover design. Cloud-agnostic. |
Availability is bought, not wished for. Vendor primitives from memory-long §stack; the economics and discipline are universal.
RTO (how long down) and RPO (how much data lost) are business requirements with price tags — the human sets them via ALIGN, architecture follows. Designing HA without stated RTO/RPO is over- or under-engineering by luck.
| Tier | RTO | Cost |
|---|---|---|
| Backup & restore | hours-days | $ |
| Pilot light (data replicated, compute off) | tens of minutes | $$ |
| Warm standby (scaled-down live copy) | minutes | $$$ |
| Active-active (multi-region) | ~zero | $$$$ + complexity tax |
Multi-AZ is the floor for production. Multi-region is a cost/complexity decision that goes through architect + fin-ops with numbers.
A backup that has never been restored does not exist. Restore drills are scheduled work (calendar, not intention) · backups live in a different failure domain than the source (account/region) · retention matches compliance, verified by policy · state files and databases both count.
Health checks probe real dependency chains (DB query, not TCP port) · failover automated, failback manual (flapping is worse than a controlled recovery) · timeouts/retries/circuit breakers at every external dependency — map what happens when each dependency dies · single-instance anything in prod = documented, accepted risk with an owner.
Every resilience claim has a test: kill an instance, drop an AZ (game day when maturity allows) · DR runbook exists, is versioned, and names humans · report residual risk honestly (what we do NOT survive).
