subagent-driven-development

Dispatch fresh subagents per task and run two-stage spec and code quality reviews.

1|Updated Jan 17, 2026
One-click install
npx skills add https://github.com/RawToast/OwO --skill subagent-driven-development-rawtoast
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/RawToast/OwO/tree/main/packages/example/skill/subagent-driven-development
Command: npx skills add https://github.com/RawToast/OwO --skill subagent-driven-development-rawtoast

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables executing complex implementation plans by dispatching dedicated subagents for each task within the same session, improving focus and speed through isolated execution and iterative reviews.

Core Features & Use Cases

  • Two-stage review workflow: spec compliance followed by code quality review for every task to ensure correctness.
  • Per-task subagents: Fresh subagents are created per task to avoid context pollution and enable parallel or sequential workflows as needed.
  • Prompt-driven orchestration: Utilizes predefined prompts for implementer, spec reviewer, and code quality reviewer to standardize interactions.

Quick Start

Invoke a plan executor to dispatch a fresh implementer subagent for each task and run the two-stage review workflow.

Frequently Asked Questions about subagent-driven-development

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I orchestrate subagents to execute complex implementation plans?

You orchestrate subagents by using a controller to dispatch a fresh implementer for each plan task within a single session. This isolates execution context and runs a two-stage review workflow to ensure spec compliance and code quality.

What is a two-stage review workflow for code generation?

A two-stage review workflow for code generation is a process that checks each task sequentially for spec compliance and code quality. It uses dedicated reviewer subagents to standardize interactions and allow iterative refinement after every implementation step.

How do I prevent context pollution when running multi-step coding tasks?

You prevent context pollution in multi-step coding tasks by creating a fresh subagent for each individual task. This isolated execution environment ensures that previous task contexts do not interfere with subsequent implementations or reviews.

Can I use prompt pipelines to standardize code review and spec compliance?

Yes, you can standardize code review and spec compliance using predefined prompt pipelines. The system relies on dedicated prompts for the implementer, spec reviewer, and code quality reviewer to orchestrate the flow and standardize all subagent interactions.

Does subagent-driven development work without external dependencies?

Yes, subagent-driven development works without external dependencies. It operates entirely within a single session using predefined prompts and a controller to orchestrate the flow between independent implementer and reviewer subagents.