# subagent-driven-development

Runs an implementation plan through fresh subagents and puts code review between tasks, so context stays clean and issues surface early.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 1300 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/subagent-driven-development/SKILL.md
- Updated: 2026-08-26
- HTML: https://affpapa.org/en/nevibecode/skills/neolabhq-subagent-driven-development

## What it does

The skill takes a plan file and executes it with subagents: one agent per task or issue, then a review of the code and output after each task or batch. The parent session keeps its context clean because every subagent starts fresh. No human sits between tasks, which is where the speed comes from.

Two modes. Sequential is for tasks that depend on each other: one agent per task, review after each one. Parallel is for unrelated work — different files, subsystems, bugs — with one agent per problem domain and review after a batch. For batches the skill advises starting with the first three tasks, then reporting back and waiting for feedback.

Red flags are spelled out: never skip review between tasks, never dispatch several implementation subagents in parallel because they conflict, and never fix a failed task by hand — that is what a fix subagent is for. Hit a blocker mid-batch and the skill tells you to stop and ask instead of guessing. Finishing the work requires the `superpowers:finishing-a-development-branch` sub-skill; without it the last step does not run.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill subagent-driven-development -a claude-code
```

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

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill subagent-driven-development -a codex
```

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

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill subagent-driven-development -a cursor
```

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

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install NeoLabHQ/context-engineering-kit subagent-driven-development
```

2. **Manually: put the files in place**
- File: `.github/skills/subagent-driven-development/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill subagent-driven-development -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/NeoLabHQ/context-engineering-kit --path skills/subagent-driven-development
```

### Windsurf (Devin Desktop)

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

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add NeoLabHQ/context-engineering-kit --skill subagent-driven-development -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/subagent-driven-development/SKILL.md

## When to use it
- execute a plan of independent tasks without polluting session context
- review code after every task instead of at the end
- run several unrelated bugs in parallel across subsystems
- fix review findings with a dedicated subagent
- close out the branch through `finishing-a-development-branch`

## Example prompts
- Use subagent-driven-development to execute the plan in `plan.md`
- Run these three independent bugs through subagent-driven-development in parallel
- Execute the plan with subagent-driven-development and review after each task

## Similar
- [doubt-driven-development](https://affpapa.org/en/nevibecode/skills/doubt-driven-development.md) — skill · 56700 installs
- [test-driven-development](https://affpapa.org/en/nevibecode/skills/test-driven-development.md) — skill · 259292 installs
- [source-driven-development](https://affpapa.org/en/nevibecode/skills/source-driven-development.md) — skill · 58200 installs
- [content-driven-development](https://affpapa.org/en/nevibecode/skills/content-driven-development.md) — skill · 1600 installs
- [context-driven-development](https://affpapa.org/en/nevibecode/skills/context-driven-development.md) — skill · 11099 installs
- [hate](https://affpapa.org/en/nevibecode/skills/hate.md) — skill · 1300 installs

## Other versions with the same name
- [subagent-driven-development · obra](https://affpapa.org/en/nevibecode/skills/subagent-driven-development.md) — 237746 installs
- [subagent-driven-development · ganyuanran](https://affpapa.org/en/nevibecode/skills/ganyuanran-subagent-driven-development.md) — ★ 1384

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