Can subagents communicate directly with one another?

No, all communication is mediated by Claude. Claude serves as the information bridge, converting output from one agent into the structured input required for the next.

How does autonomous execution mode work?

After a work plan is approved, Claude enters a mode where subagents decompose tasks, implement code, and run quality fixes automatically without requiring human approval at every step.

What happens if a requirement change is detected mid-flow?

The system features a change detection checklist; if new features or constraints are identified, autonomous execution stops and the workflow returns to the requirement analysis phase.

What is the primary role of Claude when using this skill?

Claude acts as a central orchestrator rather than a direct executor, managing specialized subagents that handle specific phases like analysis, design, coding, and quality assurance.

What triggers a project scale determination?

The requirement-analyzer subagent assesses the task and determines scale (Small, Medium, or Large) based on the estimated number of files affected and technical complexity.

Subagent Orchestration Guide

Name: Subagent Orchestration Guide
Author: shinpr

byshinpr

•

142

生産性とワークフロー

Orchestrates specialized AI subagents to manage complex coding workflows through structured requirement analysis and autonomous implementation.

概要

This skill transforms Claude into a sophisticated project orchestrator that coordinates a fleet of specialized subagents for end-to-end software development. By enforcing a scale-based workflow, it ensures that every task—from minor fixes to major feature additions—follows a rigorous path through requirement analysis, PRD creation, technical design, and automated testing. It is specifically designed for high-stakes production environments where quality assurance, structured documentation, and autonomous execution of implementation tasks are critical for project success.

主な機能

Autonomous execution mode for implementation, testing, and self-healing quality fixes
Integrated requirement change detection to prevent project drift
Scale-based workflow determination for Small, Medium, and Large projects
Automated coordination between specialized documentation and execution subagents
Standardized approval gates for PRDs, ADRs, and technical design documents
142 GitHub stars

ユースケース

Automating the generation and review of technical documentation like PRDs and Design Docs
Managing autonomous coding cycles including task decomposition and git commit execution
Orchestrating complex feature additions across multiple files with automated quality assurance

概要

主な機能

Autonomous execution mode for implementation, testing, and self-healing quality fixes
Integrated requirement change detection to prevent project drift
Scale-based workflow determination for Small, Medium, and Large projects
Automated coordination between specialized documentation and execution subagents
Standardized approval gates for PRDs, ADRs, and technical design documents
142 GitHub stars

ユースケース

Automating the generation and review of technical documentation like PRDs and Design Docs
Managing autonomous coding cycles including task decomposition and git commit execution
Orchestrating complex feature additions across multiple files with automated quality assurance