Verified:GitHub

subagent-driven-development

Delegate coding tasks to isolated subagents with structured reviews.

15|Updated Feb 14, 2026

Search Tags:#testing #code review #software development #subagents #task delegation #implementation plans #specification compliance

Ready to run

npx skills add https://github.com/mahoushoujyo-eee/eshell --skill subagent-driven-development-mahoushoujyo-eee

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/mahoushoujyo-eee/eshell/tree/main/.codex/skills/subagent-driven-development
Command: npx skills add https://github.com/mahoushoujyo-eee/eshell --skill subagent-driven-development-mahoushoujyo-eee

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents implementation plans from becoming inconsistent or incomplete by coordinating isolated subagents and enforcing structured review gates for every task.

Core Features & Use Cases

Task Delegation: Dispatch a fresh subagent for each independent implementation task with focused context and responsibilities.
Two-Stage Reviews: Require specification compliance review before code quality review, with fixes and re-review cycles for identified issues.
Continuous Execution: Complete the full plan without unnecessary pauses while handling questions, blockers, testing, and final review systematically.
Use Case: Apply it to a multi-task software feature plan where each change needs implementation, testing, specification validation, and maintainability checks before completion.

Quick Start

Use the subagent-driven-development skill to execute the implementation plan one task at a time with independent implementers and mandatory specification and code quality reviews.

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 multi-task software implementation plans without losing context?▼

Multi-task software implementation plans are executed by dispatching independent coding tasks to isolated subagents, maintaining focused context for each task. This approach ensures continuous execution while systematically handling testing, blockers, and structured reviews for every change.

How does subagent task delegation work for software development?▼

Subagent task delegation works by dispatching a fresh, isolated subagent for each independent implementation task with focused context. Each completed task undergoes mandatory specification compliance and code quality reviews before integration into the overall plan.

What is the best way to enforce specification compliance during code reviews?▼

The best way to enforce specification compliance is through a two-stage review process. A specification compliance review is required before a code quality review, with mandatory fix and re-review cycles for any identified issues before proceeding.

Can I use isolated subagents for multi-task software feature development?▼

Yes, isolated subagents are ideal for multi-task software feature development requiring focused context. Each change is handled by a fresh subagent and undergoes implementation, testing, specification validation, and maintainability checks before completion.

Why does my implementation plan become inconsistent across multiple tasks?▼

Implementation plans become inconsistent without coordinating isolated subagents and enforcing structured review gates. Applying continuous execution with mandatory specification compliance and code quality reviews prevents plans from becoming incomplete or inconsistent.

When do I need structured review gates for task delegation?▼

Structured review gates are needed for task delegation when executing multi-task software development plans where each change requires implementation, testing, specification validation, and maintainability checks before completion to ensure final integration quality.