← ClaudeAtlas

prompt-reviewlisted

Audits a candidate prompt (.prompt.md) against five gates (spec compliance, content quality, scope fit, safety, currency & coherence). Use when reviewing a new prompt draft before commit, evaluating a Mall prompt or store prompt for adoption, or re-auditing existing prompts on a periodic cadence.
fabioc-aloha/Alex_ACT_Edition · ★ 0 · AI & Automation · score 73
Install: claude install-skill fabioc-aloha/Alex_ACT_Edition
<!-- intentional divergence from Supervisor: Edition expands gate names inline and strips Supervisor-only ADR-007 ref. Same five gates. Audited 2026-05-31. --> # Prompt Review Audit a candidate prompt against the five-gate contract. Prompts are workflow entry points — short, imperative, single-goal. No optional Gate 6 (prompts are short by nature; no scaling concern Gates 1–5 miss). ## When to Use Three fire contexts: 1. **Author self-audit** — dogfood your own draft before committing. Invoked from [prompt-creator](../prompt-creator/SKILL.md) Phase 7. 2. **External candidate adoption** — gate before pulling a Mall prompt or store prompt into this brain. 3. **Periodic re-audit** — re-check existing prompts on a cadence (e.g., quarterly retraining). The shared gate model lives in [skill-review/SKILL.md § The Five Gates](../skill-review/SKILL.md). This file documents the **prompt-specific criteria** for each gate. A mechanical validator (e.g., a brain-qa script that ships with the heir) typically checks Gate 1 frontmatter compliance and the mechanical subset of Gate 5. Gates 2–5 are judgment-only. ## The Five Gates A candidate must pass **all five** to ship. Failure on any gate = decline or revise. ### Gate 1 — Spec Compliance | Check | Pass criterion | |---|---| | Frontmatter present AND minimal | `description` + `lastReviewed`. Reject any of: `name`, `type`, `application`, `tier`, `currency`, `inheritance`, `lifecycle`, `user-invokable`, `evidence`. | | `mode:` fiel