hamen

PR Review Relay — AI skill for Claude Code

AI community

Hand a PR off to your other AI coding agents (Claude/Codex/Cursor) for automated, free, local cross-review.

How to install PR Review Relay

This entry records only its repository, not the path inside it, so there is no exact command to give. Open hamen/pr-review-relay and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What PR Review Relay does

Hand a PR off to your other AI coding agents (Claude/Codex/Cursor) for automated, free, local cross-review.

Alternatives in AI

Codeburn — Free, local tool to track AI coding token usage and cost across 37 tools and agents (Claude Code, Cursor, Code 9.7k ★
Codex — openai-codex Adversarial code review, Codex CLI integration, cross-model analysis 9.6k ★
Token Monitor — Local-first desktop widget for tracking token usage, costs, and limits across 32+ AI coding tools—including Cl 1.8k ★

README

pr-review-relay

![header](assets/header.png)

[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE) [![Shell Script](https://img.shields.io/badge/shell-bash-89e051?logo=gnu-bash&logoColor=white)](pr-review-relay) [![Works with Claude](https://img.shields.io/badge/works%20with-Claude%20Code-blueviolet?logo=anthropic&logoColor=white)](https://docs.anthropic.com/en/docs/claude-code) [![Works with Codex](https://img.shields.io/badge/works%20with-Codex%20CLI-green?logo=openai&logoColor=white)](https://github.com/openai/codex) [![Works with Cursor](https://img.shields.io/badge/works%20with-Cursor-0098FF?logo=cursor&logoColor=white)](https://cursor.com) [![Works with Antigravity](https://img.shields.io/badge/works%20with-Antigravity-orange)](https://antigravity.dev) [![Works with OpenCode](https://img.shields.io/badge/works%20with-OpenCode-white)](https://opencode.ai) [![Works with Qwen Code](https://img.shields.io/badge/works%20with-Qwen%20Code-yellow)](https://qwen.ai/qwencode)

**Hand a pull request off to your *other* AI coding agents for an automated cross-review.**

You build a feature with one agent (Claude Code, Codex, Cursor, or Antigravity), it opens a PR — and the **others** automatically review that PR, headless, and post their findings as PR comments. (Reviewers are *asked* to be read-only; OpenCode and Grok enforce read-only (Grok via --deny '*' + sandbox; OpenCode via its own agent policy) — see [Notes & caveats](#-notes--caveats).) Local, free (it uses the agent CLIs you already pay for), and idempotent.

 build feature  ──►  open PR  ──►  pr-review-relay --author 
                                         │
         ┌───────────────────────────────┼───────────────────────────────┐
         ▼                               ▼                               ▼
   claude -p                       codex exec                cursor-agent -p --model 
   agy -p          opencode --pur