Claude Review Loop

review-loop

A Claude Code plugin that adds an automated code review loop to your workflow.

What it does

When you use /review-loop, the plugin creates a two-phase lifecycle:

1. Task phase: You describe a task, Claude implements it
2. Review phase: When Claude finishes, the stop hook prepares a Codex runner script and blocks exit. Claude then runs Codex directly (with output streaming to the user) and addresses the review feedback.

The result: every task gets an independent second opinion before you accept the changes, and you can watch the review happen in real time.

Review coverage

The plugin spawns up to 4 parallel Codex sub-agents, depending on project type:

After all agents finish, Codex deduplicates findings and writes a single consolidated review to reviews/review-.md.

Requirements

- Claude Code (CLI)
- jq — brew install jq (macOS) / apt install jq (Linux)
- Codex CLI — npm install -g @openai/codex

Codex multi-agent

This plugin uses Codex multi-agent to run parallel review agents. The /review-loop command automatically enables it in ~/.codex/config.toml on first use.

To set it up manually instead:

Installation

From the CLI:

Or from within a Claude Code session:

Updating

Usage

Start a review loop

Claude will implement the task. When it finishes, the stop hook:
1. Prepares a Codex runner script and prompt file
2. Blocks Claude's exit with instructions to run the review
3. Claude runs bash .claude/review-loop-run-codex.sh — Codex output streams to the user
4. Codex writes findings to reviews/review-.md
5. Claude reads the review, addresses items it agrees with, then stops

Cancel a review loop

How it works

The plugin uses a Stop hook — Claude Code's mechanism for intercepting agent exit. When Claude tries to stop:

1. The hook reads the state…