What are the built-in agents available in Claude Code?

Claude Code includes several built-in agents such as Explore for codebase search, Plan for implementation design, and a general-purpose agent for multi-step tasks.

What is the difference between a Claude Code Agent and a Skill?

Agents run in separate, isolated sessions with autonomous execution and full tool access, whereas Skills provide knowledge or guidance context within the main session.

Where do I store my custom agent definitions?

Custom agents are stored as Markdown files in .claude/agents/ for project-specific tasks or in ~/.claude/agents/ for global use across all your projects.

Can Claude invoke subagents automatically?

Yes, Claude can intelligently decide to invoke a specific subagent based on the user's request and the task context, such as calling a security reviewer for authentication code.

How can I monitor what a subagent has changed?

You can use the SubagentStop hook to programmatically analyze transcripts, track file edits, and log activity once an agent completes its task.

Claude Code Agents

Name: Claude Code Agents
Author: constellos

byconstellos

•

Productivity & Workflow

Enables the creation and orchestration of specialized subagents for complex, autonomous multi-step development tasks.

Claude Code Agents provides a robust framework for defining and deploying specialized AI subagents that operate in isolated sessions with their own unique context and toolsets. By utilizing standardized Markdown definitions, developers can create domain-specific assistants—such as security auditors, performance optimizers, or test generators—that Claude can invoke automatically or manually to handle intricate workflows. This system allows for deeper codebase exploration and focused execution of background tasks while keeping the main session transcript clean and manageable.

Key Features

01Lifecycle hooks for tracking subagent start and stop events

02Automatic and manual subagent invocation capabilities

03Support for project-specific, global, and plugin-based agents

04Markdown-based agent definitions with YAML metadata

053 GitHub stars

06Isolated execution environments with separate session context

Use Cases

01Automating multi-step security audits and vulnerability scanning

02Creating specialized assistants for complex codebase refactoring or migrations

03Generating comprehensive implementation plans before starting large features

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add constellos/claude-code-plugins claude-agents

For use in Claude.ai and ChatGPT

Key Features

01Lifecycle hooks for tracking subagent start and stop events

02Automatic and manual subagent invocation capabilities

03Support for project-specific, global, and plugin-based agents

04Markdown-based agent definitions with YAML metadata

053 GitHub stars

06Isolated execution environments with separate session context

Use Cases

01Automating multi-step security audits and vulnerability scanning

02Creating specialized assistants for complex codebase refactoring or migrations

03Generating comprehensive implementation plans before starting large features

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add constellos/claude-code-plugins claude-agents

For use in Claude.ai and ChatGPT