메뉴
Murphy-law hat panel for unhappy paths -- input, timing, system, state, UX-recovery hats probe a change in parallel. Use when forms, validation, async/data, mutations, cache, state machines, destructive actions, or loading/empty/error states change.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Router for Ben's frontend-skills harness and every local skill.
Evidence-backed Go review for bounds, APIs, concurrency, errors, security, tests, rollout, and controllers. Use when reviewing a Go diff, PR, or branch, or as the golang hat /review adds when the diff touches *.go, go.mod, or backend protos.
Evidence-backed Go conventions for workload bounds, error boundaries, proto/API implementation, Temporal, testing, rollout, and controllers. Use when writing or changing Go services, handlers, workflows, controllers, or Go tests.
Evidence-backed Go review for bounds, APIs, concurrency, errors, security, tests, rollout, and controllers. Use when reviewing a Go diff, PR, or branch, or as the golang hat /review adds when the diff touches *.go, go.mod, or backend protos.
Evidence-backed Go conventions for workload bounds, error boundaries, proto/API implementation, Temporal, testing, rollout, and controllers. Use when writing or changing Go services, handlers, workflows, controllers, or Go tests.
Reviews a diff with a 7-hat parallel panel -- product/spec, standards, complexity, adversarial, resilience, visual/design, test/perf -- plus a mandatory cross-family GPT hat. Use when reviewing a branch, PR, WIP, or a release audit (deep mode).
| name | resilience-review |
| description | Murphy-law hat panel for unhappy paths -- input, timing, system, state, UX-recovery hats probe a change in parallel. Use when forms, validation, async/data, mutations, cache, state machines, destructive actions, or loading/empty/error states change. |
Murphy pass: find every plausible unhappy path, then block, guard, recover, observe.
Diffs touching forms, validation, submit, async/data, mutations, cache, retries, state machines, mode/union switches, config/resource choice, destructive actions, or loading/empty/error/success UI. Skip docs/test/style/trivial pure logic only; record reason.
Map the risk surface first (user action, path, state change, side effects, deps). Claude-hosted
sessions may spawn the hats in parallel. Native Codex runs every axis inline unless the user
explicitly requests agents or invokes /swarm; skill activation alone is not consent. Each hat
owns one failure class with non-goals, so findings do not converge. Small diffs stay inline everywhere.
| Hat | Owns (and probes) | Non-goals |
|---|---|---|
| input | empty, null, duplicate, malformed, stale, huge, out-of-order payloads; validation format-vs-presence | timing, infra |
| timing | double submit, tab race, retry storms, slow network, timeout, cancel, out-of-order responses | payload validity |
| system | partial outage, 500s, stale cache, deleted resource, permission drift, dependency failure | UI polish |
| state | invalid/impossible states, mode-switch residue, oneof cleanup, stale derived state, recovery to a good state | network causes |
| ux-recovery | unclear disabled states, lost errors, fake success, dead ends without retry/undo, missing loading/empty states | root causes owned by other hats |
Each hat emits findings with: scenario, trigger, expected behavior, guard (Precondition -> Postcondition -> Fallback -> Observability), test to write, evidence (file/route/form/API cited). For external/browser/platform behavior, the hat runs /read-the-damn-docs; complex planned state flows sketch /visual-plan first.
Merge: dedupe by root cause, keep the highest-severity framing, then drive the finding loop: /diagnosing-bugs feedback loop -> /tdd RED test/snapshot -> /visual-review for UI validation.
## Resilience review
Risk surface:
- ...
Hats: input <status> | timing <status> | system <status> | state <status> | ux-recovery <status>
Failure matrix:
| Scenario | Trigger | Expected behavior | Guard | Test | Observability |
Finding queue:
| Finding | Repro/diagnosing-bugs loop | RED test or snapshot | Owner | Visual review needed |
Required tests:
- ...
Polish gaps:
- ...
Verdict: PASS | NEEDS_GUARDS | BLOCKED
Rules: cite files/routes/forms/API. Docs/help text not enough when code can prevent bad state. Real gap unfixed -> PR evidence with owner/deferral. No silent hat skips -- one-line evidence per skipped hat.
See REFERENCE.md.