skills.rest
cloudbro-kube-ai
cloudbro-kube-aiOFFICIAL
Explore all skills by this author →
GitHub

subagent-driven-development

Dispatch fresh subagents for isolated task execution with two-stage reviews.

14|5|Updated Jan 25, 2026
Search Tags:#workflow#tdd#subagents#code-review#ai-development#task-dispatch#spec-review
One-click install
npx skills add https://github.com/cloudbro-kube-ai/k13d --skill subagent-driven-development-cloudbro-kube-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/cloudbro-kube-ai/k13d/tree/main/skills/superpowers/subagent-driven-development
Command: npx skills add https://github.com/cloudbro-kube-ai/k13d --skill subagent-driven-development-cloudbro-kube-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Subagent-driven development helps teams execute complex implementation plans by distributing work across fresh subagents, preventing context leakage and enabling structured two-stage reviews.

Core Features & Use Cases

  • Fresh subagents per task with isolated context to improve focus and reduce cross-task interference.
  • Two-stage reviews: spec compliance first, then code quality, to catch issues early.
  • Structured task orchestration within a single session for fast iteration and visibility.

Quick Start

Start a per-task subagent workflow for the current plan and run the two-stage review (spec compliance first, then code quality) 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 prevent context leakage when executing multiple AI development tasks in one session?▼

Dispatching fresh subagents for each task ensures isolated context and prevents cross-task interference during AI development. This approach maintains clean subagent context for fast, focused task execution.

What is two-stage code review and how does it work with subagents?▼

Two-stage code review checks spec compliance first, then evaluates code quality. This structured workflow uses dedicated reviewer subagents to catch issues early before final consolidation.

How to run parallel implementation tasks while keeping spec review checkpoints?▼

Dispatch independent subagents to execute multiple tasks in parallel within the same session. This enables fast iteration while preserving structured spec compliance and code quality review checkpoints.

Can I use subagent-driven development for plans with dependent tasks?▼

Subagent-driven development is applicable to plans with multiple independent tasks. Dependent tasks requiring sequential execution are not ideal for this isolated subagent dispatch workflow.

Why does my AI coding agent lose focus when processing a large implementation plan?▼

Large plans cause context leakage and cross-task interference. Using fresh subagents per task isolates context, improving focus and preventing the agent from losing track of the overall spec.

Does subagent-driven development require external dependencies or components?▼

Subagent-driven development requires no external dependencies or components. It operates entirely within your session using built-in subagent dispatch and task orchestration mechanisms.

Related Skills

Explore More Software Engineering Skills→
haojing8312haojing8312
154

subagent-driven-development

Execute implementation plans by dispatching fresh subagents for each independent task.

Community
Intermediate
WesleyMFrederickWesleyMFrederick

merging-feature-branches-to-main

Squashes Git feature-branch history into a single commit when merging to main.

Community
Basic
bobmatnycbobmatnyc
146

Requesting Code Review

Dispatch a code-reviewer subagent to review changes against plans and requirements.

Community
Basic
royhenengelroyhenengel

run-prompt

Execute prompts from ./prompts/ in isolated sub-task contexts.

Community
Intermediate