Command Development

Create a slash command by writing a Markdown file in `.claude/commands/` with YAML frontmatter defining metadata, then arguments and bash execution blocks. Use the `!` syntax for bash execution and `@` syntax for file references to build dynamic, reusable commands within Claude Code workflows.

Slash commands use Markdown format with YAML frontmatter at the top specifying command name and parameters, followed by the command logic. Frontmatter defines dynamic arguments, file inputs, and execution context; the body contains instructions and bash execution blocks for automation.

Yes. Use `@` syntax to reference files and `!` syntax to execute bash commands within your slash commands. This enables dynamic context retrieval and system-level automation, letting you build commands that integrate file analysis, testing, and deployment workflows.

Leverage the `${CLAUDE_PLUGIN_ROOT}` variable in your command's bash execution blocks to access plugin resources portably. This enables commands to reference plugin scripts and assets without hardcoded paths, making commands work across different plugin installations.

Validate arguments in your command's frontmatter by specifying required parameters and their types. Use bash execution blocks to validate input data, check file existence, and handle edge cases before proceeding with the core command logic.

Yes. Create separate `.md` files in `.claude/commands/` for each command. Each file is independently discoverable and executable, allowing you to organize related commands—like deployment, testing, and review workflows—as a cohesive plugin suite.