Claude Code Learning Project

A learning project for exploring Claude Code capabilities, specifically Subagents and Skills.

Overview

This project demonstrates the use of specialized AI assistants (subagents) and custom capabilities (skills) within Claude Code. It serves as both a learning resource and a practical example of how to leverage these features.

What You'll Learn

1. Subagents - Specialized AI assistants with custom prompts, tools, and permissions
2. Skills - Custom capabilities you create as folders with instructions

Project Structure

subagents-test/
├── .claude/
│   ├── agents/              # Project-specific subagents
│   └── skills/              # Project-specific skills
├── src/
│   ├── components/          # React components
│   ├── services/            # API services
│   └── utils/               # Utility functions
├── PLAN.md                  # Detailed learning plan and reference
└── README.md                # This file

Subagents

Subagents are specialized AI assistants that:

• Run in their own context window
• Have custom system prompts
• Have specific tool access (allowlist/denylist)
• Have independent permission modes

Available Subagents

Subagent	Purpose	Tools Available
Bash	Command execution, git operations, terminal tasks	Bash
Explore	Fast codebase exploration, finding files, searching	All tools except Task, ExitPlanMode, Edit, Write, NotebookEdit
Plan	Software architecture, implementation planning	All tools except Task, ExitPlanMode, Edit, Write, NotebookEdit
general-purpose	Complex research, multi-step tasks, searching code	All tools (*)
claude-code-guide	Questions about Claude Code CLI, Agent SDK, hooks, MCP	Glob, Grep, Read, WebFetch, WebSearch
statusline-setup	Configure Claude Code status line setting	Read, Edit

Creating Custom Subagents

Create a .md file in .claude/agents/ with YAML frontmatter:

---
name: my-agent
description: When to use this agent
tools: Read, Grep, Glob
model: sonnet
---

You are a specialized assistant. Your task is...

Frontmatter Fields

Field	Required	Description
name	Yes	Unique identifier (lowercase, hyphens)
description	Yes	When Claude should delegate to this agent
tools	No	Tools agent can use (inherits all if omitted)
disallowedTools	No	Tools to deny
model	No	sonnet, opus, haiku, or inherit
permissionMode	No	default, acceptEdits, dontAsk, bypassPermissions, plan
skills	No	Skills to preload into agent context
hooks	No	Lifecycle hooks scoped to this agent

Skills

Skills are folders with instructions that teach Claude specific tasks. They live in:

.claude/skills/your-skill-name/skill.md

Skill Structure

---
name: skill-name
description: Brief description of what this skill does
---

# Skill Title

Detailed instructions for Claude on how to handle this skill.

Skills Paths by Tool

Tool	Project Path	Global Path
Claude Code	.claude/skills/	~/.claude/skills/
Cursor	.cursor/skills/	~/.cursor/skills/
Codex	.codex/skills/	~/.codex/skills/

Popular Official Skills

From the awesome-agent-skills repo:

By Anthropic:

• anthropics/pdf - PDF manipulation
• anthropics/mcp-builder - Create MCP servers
• anthropics/skill-creator - Guide for creating skills

By Vercel:

• vercel-labs/react-best-practices - React patterns
• vercel-labs/next-best-practices - Next.js patterns

By Trail of Bits (Security):

• trailofbits/audit-context-building - Deep code analysis
• trailofbits/static-analysis - CodeQL, Semgrep

Source Code

The project includes example TypeScript/React code:

• src/components/Button.tsx - A reusable React button component with variants
• src/services/api.ts - Generic API fetch wrapper with TypeScript types
• src/utils/helpers.ts - Utility functions (formatDate, debounce)

Examples to Try

1. Explore Agent - Quick Search

Find all TypeScript files with "interface" keyword

2. General Purpose - Research

Research best practices for error handling in Node.js

3. Plan Agent - Architecture

Design a REST API for a todo application

4. Parallel Agents

Run multiple agents simultaneously by sending multiple tool calls in a single message.

Progress

• Create plan file
• Run Explore agent example (found 2 interfaces)
• Demonstrate parallel subagents (3 agents at once)
• Run general-purpose agent example (code pattern analysis)
• Run Plan agent example
• Experiment with background agents
• Try resuming an agent

Resources

• Claude Code Documentation
• awesome-agent-skills - Community-maintained list of agent skills
• PLAN.md - Detailed learning plan and reference guide

License

MIT