What is a Claude Code subagent?

A subagent is a specialized Claude instance that runs in an isolated context with a focused role and specific tool access to perform autonomous tasks.

How do I run multiple agents at once?

You can use the Task tool with the 'run_in_background' parameter set to true. This allows you to launch several agents simultaneously and retrieve their results later using TaskOutput.

Can subagents ask me questions while they run?

No, subagents are designed to be non-interactive 'black boxes.' They cannot use tools like AskUserQuestion and must return their final output to the main conversation.

Why should I use XML tags instead of Markdown headings in prompts?

XML tags provide a clearer semantic structure for Claude to parse, which improves token efficiency and ensures the model follows complex role instructions more accurately.

Subagent & Workflow Creator

Name: Subagent & Workflow Creator
Author: Melvynx

byMelvynx

•

128

•

生産性とワークフロー

Automates the creation, configuration, and orchestration of specialized Claude Code subagents for complex multi-agent developer workflows.

This skill provides a comprehensive framework for designing, configuring, and managing specialized Claude Code subagents that operate in isolated contexts. It guides users through the entire lifecycle of agent creation—from defining lowercase-with-hyphens naming conventions to writing high-performance, XML-structured system prompts. By mastering the Task tool, developers can implement advanced delegation patterns, including background execution and parallel processing, allowing Claude to perform autonomous research, analysis, and code generation without requiring constant user interaction.

主な機能

01Parallel agent orchestration for complex multi-tasking operations

02Standardized subagent creation workflow via the /agents command

03Background execution management using the Task and TaskOutput tools

04Support for project-level and user-level agent configurations

05Optimized XML-based system prompt engineering for better token efficiency

06128 GitHub stars

ユースケース

01Setting up a parallel analysis pipeline for security and code quality audits

02Creating autonomous researchers to crawl documentation while you code

03Building specialized 'Code Reviewer' agents for specific project standards

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add melvynx/aiblueprint meta-subagent-creator

For use in Claude.ai and ChatGPT

Download Skill