# launch-sub-agent

Dispatches a focused sub-agent with isolated context, automated model routing, and mandatory self-critique verification.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 1200 installs · ★ 1787
- Author: NeoLabHQ
- License: GPL-3.0 · policy=link-only · description only; the file stays with the author
- Source: https://github.com/NeoLabHQ/context-engineering-kit/tree/HEAD/skills/launch-sub-agent/SKILL.md
- Updated: 2026-08-26
- HTML: https://affpapa.org/en/nevibecode/skills/launch-sub-agent

## What it does

The skill sets up an orchestrator pattern. Instead of polluting the primary conversation with deep troubleshooting steps, it delegates execution to a separate sub-agent. Each worker starts in a clean context window.

Execution runs across four phases. First, it breaks down the task using Zero-shot Chain-of-Thought reasoning to gauge complexity and output volume. It assigns Opus for complex architectural choices, Sonnet for lengthy non-complex outputs, and Haiku for fast lookups. If domain profiles exist in plugins like `sdd`, it injects that persona directly into the sub-agent prompt.

Every worker prompt finishes with a strict self-critique loop requiring five verification questions before finishing. If specialized role definitions are missing, the skill falls back to a general sub-agent. That gives you immediate context isolation, but specialized domain guidance depends on the plugins configured in your project.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill launch-sub-agent -a claude-code
```

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

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill launch-sub-agent -a codex
```

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

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill launch-sub-agent -a cursor
```

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

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install NeoLabHQ/context-engineering-kit launch-sub-agent
```

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

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill launch-sub-agent -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/NeoLabHQ/context-engineering-kit --path skills/launch-sub-agent
```

### Windsurf (Devin Desktop)

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

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill launch-sub-agent -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/NeoLabHQ/context-engineering-kit/tree/HEAD/skills/launch-sub-agent/SKILL.md

## When to use it
- isolate task execution to prevent primary context bloat
- select Opus, Sonnet, or Haiku based on task complexity
- attach specialized agent personas to focused sub-tasks
- enforce systematic self-critique before task completion

## Example prompts
- Use launch-sub-agent to design the database architecture in an isolated session
- Run launch-sub-agent to research third-party auth libraries
- Delegate unit test generation to a sub-agent using launch-sub-agent

## Similar
- [sub-agents](https://affpapa.org/en/nevibecode/skills/sub-agents.md) — skill · 533 installs
- [research](https://affpapa.org/en/nevibecode/skills/parcadei-research.md) — skill · 485 installs
- [ai-shaped-readiness-advisor](https://affpapa.org/en/nevibecode/skills/ai-shaped-readiness-advisor.md) — skill · 2700 installs
- [init](https://affpapa.org/en/nevibecode/skills/init.md) — skill · 1300 installs
- [context-engineering](https://affpapa.org/en/nevibecode/skills/mrgoonie-context-engineering.md) — skill · 495 installs
- [context-compression](https://affpapa.org/en/nevibecode/skills/muratcankoylan-context-compression.md) — skill · 459 installs

From the same set: [neolabhq/context-engineering-kit](https://affpapa.org/en/nevibecode/by/neolabhq/context-engineering-kit.md) — 64 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