impartial-review

name

impartial-review

description

Critically review recent code changes (uncommitted diff, latest commit, or active PR) for bugs, schema drift, edge cases, observability gaps, and project-specific rule violations. Use when the user asks to review, audit, check, stress-test, or "look hard at" recent work — including code Claude itself just wrote. Dispatches five parallel Sonnet 4.6 subagents (general-purpose, fresh context) — four bucket reviewers plus one project-aware reviewer loaded with CLAUDE.md and `.claude/reference/` — so reviewers are genuinely impartial AND aware of project-specific landmines. Verifies claims with grep/read before asserting, severity-tags findings (blocking / should-fix / nitpick), surfaces things that look suspicious but verified fine, and gives concrete fixes with file:line references.

Codex Adapter

Read ../../../.claude/skills/impartial-review/SKILL.md completely and follow it as the canonical workflow for this skill.

• Resolve every relative script, reference, asset, and template path from the canonical skill directory under .claude/skills/impartial-review/.
• Read ../../../.agents/CODEX-SKILL-COMPATIBILITY.md and apply the status and exceptions for this skill.
• Translate Claude Code tool names to the available Codex equivalent as described by the repository AGENTS.md.
• Before any Claude Task, Agent, or Workflow step, confirm that the current Codex session exposes the required multi-agent tools. Do not replace independent review with self-review.
• Skill invocation alone does not authorize commit, push, PR, merge, deploy, migration, dependency installation, or another outward action unless that action is the user's explicit request.
• Follow Codex system, developer, sandbox, approval, and tool instructions when they differ from the canonical workflow.
• Keep the canonical Claude skill unchanged unless the user explicitly asks to edit the workflow itself.