# cross-agent-review

Asks an agent on a different backend to review your code, PR or plan over the repowire mesh. Useful before a merge when you want a second opinion.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · ★ 269
- Author: prassanna-ravishankar
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/prassanna-ravishankar/repowire/tree/HEAD/skills/cross-agent-review/SKILL.md
- Updated: 2026-10-25
- HTML: https://affpapa.org/en/nevibecode/skills/cross-agent-review

## What it does

The skill resolves the reviewer in three steps. An explicit backend named by the user wins. Otherwise it reads `repowire config get skills.default_reviewer_backend`, and if that comes back empty it picks any online peer running a backend other than yours, or asks you which one to use. Your own backend is never the default here, since the whole point is a review from a different model.

To see who is around and what they run, use `list_peers()` over MCP or `repowire peer list` from the CLI. The skill then finds or spawns a peer and hands it the concrete diff, PR or plan, with a note on what to check. Over MCP that call is `ask(peer_name, ...)`; it returns a `correlation_id` right away and does not block. The reviewer ends the thread with `ack(corr_id, <review>)`, and that reply lands back with you. Follow-ups go through `ask(reply_to=corr_id, ...)`.

The ask/ack lifecycle exists only in MCP. There is no CLI equivalent for a non-blocking ask — `repowire peer ask` is a synchronous test utility, not this workflow. When no peer on a different backend is online and you want one, the skill can spawn it through the `delegate` skill, but only after you agree.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add prassanna-ravishankar/repowire --skill cross-agent-review -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/cross-agent-review/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/cross-agent-review/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add prassanna-ravishankar/repowire --skill cross-agent-review -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/cross-agent-review/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add prassanna-ravishankar/repowire --skill cross-agent-review -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/cross-agent-review/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install prassanna-ravishankar/repowire cross-agent-review
```

2. **Manually: put the files in place**
- File: `.github/skills/cross-agent-review/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add prassanna-ravishankar/repowire --skill cross-agent-review -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/prassanna-ravishankar/repowire --path skills/cross-agent-review
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/cross-agent-review/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add prassanna-ravishankar/repowire --skill cross-agent-review -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/prassanna-ravishankar/repowire/tree/HEAD/skills/cross-agent-review/SKILL.md

## When to use it
- get an independent review of a diff before merging
- have a PR checked by a model other than the one that wrote it
- run a plan past a second agent
- send a follow-up question into the same review thread
- spawn a peer on another backend when none is online

## Example prompts
- Have codex review this diff for correctness and bugs
- Send this PR to any agent running a different backend
- Ask the reviewer what it meant by the race in `worker.py`

## Similar
- [cross-review](https://affpapa.org/en/nevibecode/skills/cross-review.md) — skill · 446 installs
- [caveman-review](https://affpapa.org/en/nevibecode/skills/caveman-review.md) — skill · 411800 installs
- [review-loop](https://affpapa.org/en/nevibecode/skills/review-loop.md) — skill · 104100 installs
- [thermo-nuclear-code-quality-review](https://affpapa.org/en/nevibecode/skills/thermo-nuclear-code-quality-review.md) — skill · 22200 installs
- [open-code-review](https://affpapa.org/en/nevibecode/skills/open-code-review.md) — skill · 6900 installs
- [printing-press-output-review](https://affpapa.org/en/nevibecode/skills/printing-press-output-review.md) — skill · 6400 installs

From the same set: [prassanna-ravishankar/repowire](https://affpapa.org/en/nevibecode/by/prassanna-ravishankar/repowire.md) — 3 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md