What problem does it solve?
You need to create slash commands that work reliably in Claude Code, but the correct structure, YAML frontmatter options, argument handling, and execution patterns are unclear.
Core Features & Use Cases
- Slash command authoring: Define commands as Markdown prompts with correct instruction framing and placement.
- YAML frontmatter configuration: Use fields like description, allowed-tools, model, argument-hint, and disable-model-invocation to control discovery and behavior.
- Dynamic inputs & execution: Inject arguments with $ARGUMENTS and $1/$2, include file contents with @ syntax, and gather context by running Bash with !
....
- Plugin-aware command patterns: Use ${CLAUDE_PLUGIN_ROOT} for portable paths and design commands that integrate with plugin scripts, resources, agents, and other skills.
- Validation and troubleshooting: Apply best practices for input/resource validation, error handling, and diagnosing missing commands or failing arguments.
Quick Start
Create a new command file that includes a YAML frontmatter description and an argument-hint, then write Claude instructions that use $1 for the user-provided value and (optionally) use !... for repository context.