skills.rest
overtimepog
overtimepogCOMMUNITY
Explore all skills by this author →
GitHub

subagent-driven-development

Execute implementation plans by dispatching fresh subagents per task with mandatory code reviews.

Updated Aug 27, 2026
Search Tags:#workflow#code review#AI development#subagents#iteration#task execution#quality gates
One-click install
npx skills add https://github.com/overtimepog/AgentTheo --skill subagent-driven-development-overtimepog
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/overtimepog/AgentTheo/tree/main/.claude/skills/superpowers/subagent-driven-development
Command: npx skills add https://github.com/overtimepog/AgentTheo --skill subagent-driven-development-overtimepog

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill manages complex implementation plans by breaking them into independent tasks, ensuring high quality through continuous code review and preventing context pollution by dispatching fresh subagents for each task.

Core Features & Use Cases

  • Fresh Subagent Per Task: Dispatches a new, unpolluted subagent for each task, ensuring focused execution without carrying over irrelevant context.
  • Code Review Checkpoints: Integrates mandatory code review after each task, catching issues early and maintaining quality gates throughout development.
  • Fast Iteration: Enables continuous progress and rapid iteration within the same session, reducing human-in-the-loop delays between tasks.
  • Use Case: When executing a multi-task implementation plan, dispatch a subagent for each task, review its work with a code-reviewer subagent, fix any issues, and then proceed to the next task.

Quick Start

Apply the subagent-driven-development skill to execute the feature-x-plan.md by dispatching subagents for 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 a multi-task implementation plan with code reviews between each task?▼

Subagent-driven development breaks your implementation plan into independent tasks, dispatching a fresh subagent for each one and running mandatory code reviews after every task before proceeding to the next, ensuring quality gates throughout execution.

Why dispatch a new subagent for each task instead of using one subagent for the entire plan?▼

Fresh subagents per task prevent context pollution and irrelevant information from prior tasks, ensuring focused execution and reducing hallucination or confusion that accumulates when a single agent carries forward the entire conversation history.

Can I iterate quickly on a multi-task plan within a single session?▼

Yes, subagent-driven development enables fast iteration by automating task dispatch, code review, issue resolution, and verification all within the same session, eliminating delays between human-in-the-loop review cycles.

What happens if a code review finds issues in a task's output?▼

Issues flagged during code review are resolved before moving to the next task, maintaining quality gates and preventing defects from cascading through dependent tasks in your implementation plan.

When should I use subagent-driven development versus a single AI session?▼

Use subagent-driven development when executing complex multi-task plans where isolation, mandatory reviews, and quality checkpoints matter more than continuous context; single sessions work better for tightly integrated single-task work requiring shared context.

Does this approach require a pre-written implementation plan?▼

Yes, subagent-driven development expects an implementation plan document defining independent tasks; the workflow dispatches subagents task-by-task and orchestrates reviews and verification based on that plan structure.

Related Skills

Explore More Software Engineering Skills→
raas-devraas-dev
19

requesting-code-review

Dispatch a code-reviewer subagent to verify work against requirements.

Official
Intermediate
bacocobacoco
67

openspec-change-closure

Archive completed OpenSpec changes by merging delta specs into living specifications.

Community
Advanced
the-inconvenience-storethe-inconvenience-store
1

requesting-code-review

Manage code review requests and track reviewer feedback with git and Dolt.

Community
Intermediate
m16khbm16khb

subagent-driven-development

Dispatch fresh subagents per task with code review and test gates.

Community
Advanced