# cross-agent-plan

Asks a local coding agent for an implementation plan or a design critique through the Repowire remote MCP server.

- 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/plugins/repowire-remote/skills/cross-agent-plan/SKILL.md
- Updated: 2026-10-25
- HTML: https://affpapa.org/en/nevibecode/skills/prassanna-ravishankar-cross-agent-plan

## What it does

This skill is for the moments when a second opinion helps. It hands a task to an agent on another machine and brings back that agent's plan. First `list_machines` and `list_agents` locate an agent holding context for the project. Unless the user picked a backend, the skill suggests a different one so the answer is not an echo of your own.

Then comes `ask_agent`, called with exact machine and peer IDs. The request carries the goal, the constraints, design context and the tradeoffs you care about. Planning is the whole scope here — the skill states that no implementation is authorised by this request. Keep the returned request ID and pull the answer with `get_reply`, which can wait up to 20 seconds. A pending status means the work is still running.

Read the plan, then list the choices it leaves open. Treating it as a green light to start coding is not what the skill does. When sending times out, check `list_requests` for work already in flight before retrying. One hard limit: the remote tools cannot spin up a planner session that does not exist.

## Install

### Claude Code (native platform)

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

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

### Codex

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

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

### Cursor

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

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

### GitHub Copilot

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

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

### Gemini CLI

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

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

### Windsurf (Devin Desktop)

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

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add prassanna-ravishankar/repowire --skill cross-agent-plan -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/plugins/repowire-remote/skills/cross-agent-plan/SKILL.md

## When to use it
- get an implementation plan from an agent with different context
- test a design against an outside opinion before writing code
- locate a machine and agent suited to the task
- fetch the reply to an async request by its ID
- turn a returned plan into a list of open questions

## Example prompts
- Find an agent with project context via `list_machines` and `list_agents`, then ask for a plan to refactor the billing module
- Send the goal and constraints through `ask_agent`, wait for the reply and show me the unresolved choices
- Check `list_requests` if the send hangs instead of retrying blindly

## Similar
- [ce-plan](https://affpapa.org/en/nevibecode/skills/ce-plan.md) — skill · 3700 installs
- [plan](https://affpapa.org/en/nevibecode/skills/plan.md) — skill · 2800 installs
- [ralph-plan](https://affpapa.org/en/nevibecode/skills/ralph-plan.md) — skill · 2600 installs
- [speckit-plan](https://affpapa.org/en/nevibecode/skills/speckit-plan.md) — skill · 2500 installs
- [plan-crafting](https://affpapa.org/en/nevibecode/skills/plan-crafting.md) — skill · 2400 installs
- [alan-plan](https://affpapa.org/en/nevibecode/skills/alan-plan.md) — skill · 1800 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