skills.rest
lyfe2025
lyfe2025COMMUNITY
Explore all skills by this author →
GitHub

subagent-driven-development

Dispatch each plan task to an implementer subagent with two-stage reviews.

4|2|Updated Jan 25, 2026
Search Tags:#development workflow#parallel execution#subagent#AI-assisted development#task delegation
One-click install
npx skills add https://github.com/lyfe2025/lyfes-coding-skills --skill subagent-driven-development-lyfe2025
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/lyfe2025/lyfes-coding-skills/tree/main/skills/superpowers/subagent-driven-development
Command: npx skills add https://github.com/lyfe2025/lyfes-coding-skills --skill subagent-driven-development-lyfe2025

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill addresses the challenge of coordinating complex plans with many tasks by distributing each task to a separate subagent, ensuring isolated execution and reliable reviews.

Core Features & Use Cases

  • Per-task implementer subagents: each task runs under its own subagent to prevent context bleed.
  • Two-stage reviews: after each task, perform spec-compliance review and then code-quality review.
  • Parallel execution in a single session: tasks can be processed concurrently to accelerate delivery.
  • Self-review and traceability: subagents perform self-review before handing off.

Quick Start

Load your plan, split into tasks, dispatch an implementer subagent for each task with full task text and context, then sequentially dispatch spec compliance and code quality reviewers, and finally finish with a development-branch 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 isolate context bleed when running parallel development tasks in a single session?▼

Subagent-Driven Development isolates context bleed by dispatching each task to a separate implementer subagent, ensuring isolated execution and preventing cross-task interference within the same session.

What is the best way to enforce code quality reviews during AI-assisted development?▼

Code quality reviews during AI-assisted development are enforced through a mandatory two-stage review process, sequentially dispatching spec compliance and code quality reviewers after each implementer subagent completes a task.

How do I distribute independent tasks for parallel execution to accelerate delivery?▼

You can accelerate delivery by splitting your plan into largely independent tasks and dispatching a dedicated implementer subagent for each, allowing parallel execution within a single session.

Can I ensure traceable outcomes when delegating development tasks to subagents?▼

Traceable outcomes are ensured by using per-task prompts and requiring subagents to perform self-review before handoff, maintaining clear tracking from task delegation through the final development-branch task.

When should I not use subagent-driven development for task delegation?▼

You should avoid subagent-driven development when your plan contains tightly coupled tasks with heavy dependencies, as this approach excels primarily with largely independent tasks that can run in parallel.

Related Skills

Explore More Software Engineering Skills→
Liuchun-ossLiuchun-oss
15

subagent-driven-development

Dispatch independent subagents for isolated task execution and two-stage review.

Community
Advanced
JhabbigJhabbig

subagent-driven-development

Coordinate subagent implementation tasks with two-stage spec and code quality reviews.

Community
Advanced
gztcodegztcode

requesting-code-review

Automate code review with static security scanning, quality gates, and an independent reviewer subagent.

Community
Advanced
hetsheth-droidhetsheth-droid

subagent-driven-development

Dispatch fresh subagents per task with isolated context and two-stage reviews.

Community
Advanced