How do I ensure my agent triggers at the right time?

The description field is critical; it must include specific triggering conditions and multiple blocks showing context, user requests, and assistant responses.

What is the difference between an agent and a command in Claude Code?

Agents are designed for autonomous, multi-step work and are triggered by context, while commands are typically user-initiated actions.

Which model should a Claude Code agent use?

It is recommended to use 'inherit' so the agent matches the parent model, but you can explicitly set it to sonnet, opus, or haiku for specific needs.

What naming conventions should I use for agent identifiers?

Identifiers must be 3-50 characters, use only lowercase letters, numbers, and hyphens, and must start and end with an alphanumeric character.

Can I limit the tools an agent can access?

Yes, you can use the 'tools' field in the frontmatter to restrict access to a specific array of tools, following the principle of least privilege.

Claude Code Agent Development

Name: Claude Code Agent Development
Author: Activer007

byActiver007

개발자 도구

Guides developers through the creation, structuring, and optimization of autonomous agents within Claude Code plugins.

소개

This skill provides a comprehensive framework for building autonomous subprocesses within the Claude Code ecosystem. It standardizes the development process by providing detailed patterns for YAML frontmatter configuration, triggering condition design, and expert system prompt engineering. Whether you are defining tool permissions or crafting specialized personas, this skill ensures your agents are reliable, well-documented, and capable of handling complex, multi-step tasks independently.

주요 기능

Best practices for triggering agent execution using example blocks
Standardized markdown structure with YAML frontmatter templates
0 GitHub stars
Tool permission management and model inheritance configuration
Expert system prompt design patterns for specialized personas
Validation rules for agent identifiers and metadata formatting

사용 사례

Building autonomous code review workflows that trigger on specific file patterns
Creating a specialized security analyzer agent with restricted read-only tool access
Designing complex multi-agent plugins with distinct visual identifiers and colors

소개

주요 기능

Best practices for triggering agent execution using example blocks
Standardized markdown structure with YAML frontmatter templates
0 GitHub stars
Tool permission management and model inheritance configuration
Expert system prompt design patterns for specialized personas
Validation rules for agent identifiers and metadata formatting

사용 사례

Building autonomous code review workflows that trigger on specific file patterns
Creating a specialized security analyzer agent with restricted read-only tool access
Designing complex multi-agent plugins with distinct visual identifiers and colors