メニュー
Use for cap4k verification and audit claims: static review, focused local evidence, full evidence, generated-versus-handwritten ownership checks, skipped-check disclosure, rollback targeting, and final evidence summaries.
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
Use when projecting an approved cap4k technical design into generator inputs such as DB/schema, design JSON, manifests, Gradle extension settings, addons/options, or template override decisions.
Route cap4k business-authoring work through the self-contained cap4k skill system. Use when the user asks to discover a business slice, model DDD boundaries, write cap4k technical design, author generator inputs, review plan or generated output, implement handwritten business logic, design service integration, or verify cap4k work.
Use when a coherent cap4k business brief needs tactical DDD decisions mapped to cap4k carriers, generator input expectations, rollback notes, and the next technical design phase.
Use for cap4k business service interaction design and implementation: Open Host Service entries, Published Language, external capability clients, external fact entries, inbound/outbound integration events, callbacks, message listeners, and command/query routing across service boundaries.
Use when implementing cap4k handwritten business logic inside approved generated skeletons after human generated-output review: command handlers, query handlers, subscribers, jobs, controllers, factories, specifications, domain services, Repository access, Mediator routing, and UoW persistence.
Use to review cap4k plan evidence, generated output ownership, conflict policies, template IDs, addon artifacts, generation drift, and the stop point before handwritten implementation.
| name | cap4k-verification-audit |
| description | Use for cap4k verification and audit claims: static review, focused local evidence, full evidence, generated-versus-handwritten ownership checks, skipped-check disclosure, rollback targeting, and final evidence summaries. |
Verify structure, ownership, behavior, and evidence without overstating what was checked.
../shared/rules/verification-claim-policy.mdreferences/evidence-modes.mdworkflows/run-verification-audit.mdUse when the task asks for verification, audit, final evidence, or claim
strength. Route-level routing.yaml supplies forced rollback, drift gotchas,
and runtime capability map for this route.
| Task | Read | Workflow |
|---|---|---|
| Static-only skill or docs audit | references/evidence-modes.md | workflows/run-verification-audit.md |
| Focused local evidence when allowed | references/evidence-modes.md | workflows/run-verification-audit.md |
| Final verification summary | references/evidence-modes.md | workflows/run-verification-audit.md |