What problem does it solve?
Users need to customize Claude Code's output format and verbosity to match specific tasks, whether for quick code generation, detailed explanations, or interactive learning, but lack clear guidance on available options and customization. This skill simplifies output style management, enhancing user experience.
Core Features & Use Cases
- Built-in Output Styles: Explains the "Default" (software engineering), "Explanatory" (educational insights), and "Learning" (collaborative with
TODO(human) markers) styles.
- Custom Style Creation: Guides users through creating personalized output styles, including frontmatter configuration (name, description,
keep-coding-instructions) and file locations.
- Dynamic Style Switching: Details how to use the
/output-style command to list, select, and switch between different output styles for various tasks.
- Comparison to Related Features: Clarifies the differences and appropriate use cases for output styles compared to
CLAUDE.md, append-system-prompt, agents, and slash commands.
- Use Case: A new developer wants to learn a codebase interactively. They can use this skill to switch to the "Learning" output style, which guides them with
TODO(human) markers, fostering a collaborative learn-by-doing experience.
Quick Start
Use the output-styles-meta skill to switch Claude's output to the "Explanatory" style for more detailed insights.