subagent-driven-development

Execute implementation plans by dispatching isolated subagents per task with two-stage reviews.

Updated May 5, 2026
One-click install
npx skills add https://github.com/XCmiaow/cc-deploy --skill subagent-driven-development-xcmiaow
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/XCmiaow/cc-deploy/tree/main/skills/subagent-driven-development
Command: npx skills add https://github.com/XCmiaow/cc-deploy --skill subagent-driven-development-xcmiaow

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the problem of producing high-quality implementations faster by coordinating multiple independent work items without context confusion or missing requirements.

Core Features & Use Cases

  • Fresh subagent per task: dispatches an isolated implementer with only the task-specific context so work stays focused.
  • Two-stage review checkpoints: runs a spec compliance review first, then a code quality review after each task.
  • Controlled iteration loop: re-dispatches fixes until both reviewers approve, preventing “close enough” outcomes.

Real-world use case: you have an implementation plan with 5 tasks (some are small, some touch multiple files), and you want to execute them in a single session while avoiding cross-task context pollution and ensuring correctness and maintainability.

Quick Start

Tell the controller to execute the current implementation plan using subagent-driven-development with spec review followed by code quality review after each task.

Frequently Asked Questions about subagent-driven-development

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

FAQPage Schema
How do I execute an implementation plan while preventing cross-task context pollution?

To prevent cross-task context pollution during implementation, dispatch isolated subagents per task with only task-specific context. This ensures focused execution without context inheritance from previous tasks in the current session.

What is the two-stage review workflow for subagent task orchestration?

The two-stage review workflow for subagent task orchestration runs spec compliance review first, followed by code quality review. This sequence enforces correctness before maintainability, re-dispatching fixes until both reviewers approve.

How do I enforce spec compliance and code quality when automating code generation?

To enforce spec compliance and code quality during code generation, use an iterative fix-and-re-review loop. Re-dispatch implementer subagents with specific feedback until both the spec compliance and code quality reviewers approve.

Does subagent-driven development work for dependent tasks that require context inheritance?

Subagent-driven development does not work for dependent tasks requiring context inheritance. It fits scenarios where tasks are mostly independent and must be completed within the current session using fresh implementer subagents.

What's the best way to orchestrate independent coding tasks in a single session?

The best way to orchestrate independent coding tasks in a single session is dispatching fresh subagents per task. This approach isolates work, enforces two-stage review checkpoints, and prevents close enough outcomes through controlled iteration.