Agent Skills Fundamentals

Transform how you work with AI agents by mastering the fundamentals of agent skills — reusable, portable instruction sets that turn general-purpose agents into task-specific specialists. This hands-on course takes you from understanding the structure and format of skill folders and SKILL.md files to building custom skills that solve real-world problems. You will explore how skills compare to tools, MCP, and subagents, and learn progressive disclosure techniques to manage context efficiently. Through practical exercises, you will create skills for generating practice questions from lecture notes, analyzing time series data, and building a marketing campaign analysis workflow using Anthropic's pre-built skills for Excel, PowerPoint, and Skill Creation.

Go beyond the basics by deploying your skills across the full Claude ecosystem — Claude.ai, Claude Code, the Claude API, and the Claude Agent SDK. Learn to integrate the code execution tool and files API to give Claude filesystem access and bash capabilities for running Python scripts. Build end-to-end code generation, review, and testing workflows with Claude Code, configure subagents equipped with skills for isolated specialized tasks, and create a research agent using the Claude Agent SDK that generates learning guides from documentation, GitHub repos, and web search results. Whether you are an AI developer, solutions architect, or automation specialist, this course equips you with the skills to design and deploy intelligent agent workflows that scale.