Does this skill help with versioning my agents?

Yes, it includes a recommended version tracking and changelog structure in the YAML frontmatter (Major, Minor, Patch) to help you manage the evolution of your sub-agents.

Can I restrict what tools a sub-agent can access?

Yes, the Sub-Agent Creator allows you to define specific tool patterns like Read-only, Code operations, or full Execution (Bash) to ensure the agent only performs safe, relevant actions.

Where are the generated sub-agents stored?

You can choose to store agents at the project level in .claude/agents/ for team sharing or at the user level in ~/.claude/agents/ for personal use across all projects.

What is a Claude Code sub-agent?

A sub-agent is a specialized, domain-specific AI configuration with its own system prompt and toolset designed to handle specific tasks autonomously within the Claude Code environment.

How does proactive triggering work for sub-agents?

By including specific 'PROACTIVELY' instructions and trigger phrases in the agent's description, the main Claude instance can automatically detect when to delegate a task to the sub-agent.

Sub-Agent Creator

Name: Sub-Agent Creator
Author: cskiro

bycskiro

生産性とワークフロー

Automates the generation of specialized Claude Code sub-agents with domain-specific personas, proactive triggers, and customized tool permissions.

概要

Sub-Agent Creator is a strategic utility for Claude Code designed to streamline the delegation of complex tasks by architecting focused sub-agents. It follows a rigorous 6-step workflow to define intent, persona, and instructions, ensuring every agent is optimized for token efficiency and security. Whether you need a read-only security auditor, a code-modifying debugger, or a team-shared project validator, this skill automates the creation of YAML frontmatter and system prompts following Anthropic's best practices, allowing for seamless, proactive task handoffs within your development environment.

主な機能

Support for both project-level (.claude/agents/) and global user-level storage
Token-efficient prompt optimization to reduce latency and costs
Automated YAML frontmatter generation with version tracking and changelogs
Pre-configured tool access patterns including Read-only, Code Ops, and Execution
0 GitHub stars
Structured 6-step agent architecting workflow for high-quality delegation

ユースケース

Creating a 'Security Reviewer' agent that proactively triggers on audit requests
Setting up restricted sub-agents with limited tool access for sensitive operations
Building a project-specific 'Migration Validator' shared across a development team

概要

主な機能

Support for both project-level (.claude/agents/) and global user-level storage
Token-efficient prompt optimization to reduce latency and costs
Automated YAML frontmatter generation with version tracking and changelogs
Pre-configured tool access patterns including Read-only, Code Ops, and Execution
0 GitHub stars
Structured 6-step agent architecting workflow for high-quality delegation

ユースケース

Creating a 'Security Reviewer' agent that proactively triggers on audit requests
Setting up restricted sub-agents with limited tool access for sensitive operations
Building a project-specific 'Migration Validator' shared across a development team