What problem does it solve?
This skill removes uncertainty and manual errors when authoring, updating, or troubleshooting Claude Code slash commands by providing clear rules for file placement, frontmatter, argument handling, bash execution permissions, and file references.
Core Features & Use Cases
- Command authoring guidance: Explains required frontmatter fields such as description and optional fields like allowed-tools, argument-hint, and model.
- Argument and execution patterns: Covers use of $ARGUMENTS and positional $1/$2 variables, how to include bash output safely, and allowed-tools permission patterns.
- File reference and context: Describes how to reference files with @ paths, how directory references behave, and recommendations for testing and scoping (user vs project).
- Use Case: Create a project-scoped smart commit command that gathers git status and diffs, then formats a semantic commit message using the repository context and argument hints.
Quick Start
Create a new .claude/commands/my-command.md file with a description frontmatter, add allowed-tools if running bash, and write instructions using $ARGUMENTS for capture.