subagent-driven-development

Dispatch AI subagents for implementation and multi-stage code reviews.

Updated Feb 24, 2026

Search Tags:#code review #code generation #ai development #task automation #subagent #automated testing

Ready to run

npx skills add https://github.com/nonputtipong/Oryn --skill subagent-driven-development-nonputtipong

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/nonputtipong/Oryn/tree/main/.agent/skills/subagent-driven-development
Command: npx skills add https://github.com/nonputtipong/Oryn --skill subagent-driven-development-nonputtipong

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill automates the implementation of development tasks by orchestrating specialized AI subagents, ensuring adherence to specifications and code quality standards through a structured, multi-stage review process.

Core Features & Use Cases

Task Decomposition: Breaks down implementation plans into discrete, manageable tasks.
Subagent Orchestration: Dispatches specialized AI agents for implementation, spec review, and code quality review.
Iterative Refinement: Manages feedback loops between implementer and reviewer agents to fix issues.
Use Case: When you have a detailed plan for a new feature, use this Skill to have AI agents write the code, review it against the spec, and then check for code quality, iterating until all tasks are complete and approved.

Quick Start

Use the subagent-driven-development skill to execute the implementation plan located at ./docs/plans/new-feature.md.

Frequently Asked Questions about subagent-driven-development

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

FAQPage Schema

How do I automate code generation and spec compliance for a new feature?▼

Automate code generation by dispatching independent AI subagents for implementation and multi-stage reviews. This orchestrates specialized agents to write code, verify spec compliance, and check quality iteratively until tasks are approved.

How does multi-stage code review work with AI subagents?▼

Multi-stage code review works by dispatching separate AI subagents for spec review and code quality review. The system manages iterative feedback loops between implementer and reviewer agents to fix issues until all tasks are complete and approved.

What is the best way to break down a detailed implementation plan into discrete tasks?▼

The best way to break down an implementation plan is using task decomposition via subagent orchestration. This automates the execution of discrete development tasks within a single session, managing statuses like DONE, NEEDS_CONTEXT, and BLOCKED.

How do I handle blocked tasks or missing context during AI-driven development?▼

Handle blocked tasks or missing context through automated status management that tracks NEEDS_CONTEXT and BLOCKED states. The system escalates issues as needed, managing iterative feedback loops between implementer and reviewer agents to resolve concerns.

Can I use task automation for iterative code refinement and automated testing?▼

Yes, you can use task automation for iterative code refinement. The system dispatches specialized AI agents to execute implementation, review code quality, and iterate through feedback loops, ensuring spec compliance and managing task statuses like DONE_WITH_CONCERNS.

Do I need a detailed plan to start subagent-driven development?▼

Yes, you need a detailed implementation plan to start subagent-driven development. The skill takes a structured plan, breaks it into discrete tasks, and orchestrates AI agents to execute the code generation and multi-stage review process within a single session.