Menü
Critique a PlanArtifact using a severity rubric and produce a CritiqueArtifact (Protocol-Version: 1).
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
| name | plan_critic |
| description | Critique a PlanArtifact using a severity rubric and produce a CritiqueArtifact (Protocol-Version: 1). |
The output MUST be a Markdown document:
Protocol-Version: 1## Findings## Clarifying Questions## Verification Questions## Stop SignalEach finding MUST include:
severity: Blocker|High|Medium|Lowissuewhy_it_matterssuggested_fixevidence_neededProvide 0–5 questions only if the plan would otherwise be guessing.
List concrete checks to perform in repo/config/docs to reduce risk.
Set stop_signal = true only if there are no Blocker and no High severity findings.
Verify changes against the task spec and plan. Run verify.sh, check diff quality, and produce a final compliance report.
Понимание legacy фронтенда WP_Field: SCSS, vanilla JS, npm scripts, сборка и проверка старой версии.
Понимание modern React/Vite фронтенда WP_Field: entrypoints, сборка, dev server и UI-слой.
Deterministically lint a PlanArtifact against invariants and produce a VerificationReport (Protocol-Version: 1).
Generate structured, step-by-step execution plans for coding and technical tasks. Ensures plans are atomic, testable, and ready for execution by AI agents.
Orchestrate a lightweight plan review loop (Planning → Critic → Planning → Verifier) with iteration budget and optional hooks.