managing-slash-commands

Slash commands require YAML frontmatter defining metadata, description, and argument schemas, plus execution logic. Design frontmatter to declare allowed tools, optional models, and validation steps, then structure arguments as named parameters with type hints and help text for Claude to invoke correctly.

Slash commands are direct Claude interfaces with frontmatter and arguments for structured prompts. Plugins and MCP commands extend Claude's capabilities through broader ecosystems. This Skill helps you design and document all three, managing namespaces to avoid conflicts across command types.

Use /help to validate frontmatter syntax and argument definitions, then test with /command-name to confirm execution workflows. Check that allowed-tools restrictions, argument parsing, and model selection work as designed before full deployment.

Yes. Extract the core task, structure inputs as named arguments with type hints, define frontmatter metadata including description and tools, and validate against best-practices. This Skill guides you through prompt-to-command conversion and alignment with creation guidelines.

Use clear, scoped argument names to prevent ambiguity. Apply namespace prefixes for command families, enforce allowed-tools restrictions to limit scope, and document optional parameters. Follow patterns validated against the creation guide to maintain consistency across your command ecosystem.

Allowed-tools restrictions limit which capabilities a command can invoke, reducing surface area and improving safety. They ensure commands stay focused on their intended task and prevent accidental misuse of powerful tools outside their scope.