moai-cc-commands

Slash commands in Claude Code are user-facing entry points that orchestrate multi-step workflows. Design them by defining parameters, tool permissions, and agent coordination logic, then structure them in .claude/commands with metadata, argument parsing, and validation patterns enforced.

Store commands under .claude/commands with standardized file structure including metadata headers, argument parsing rules, tool permissions declarations, and safety gates. This enforces consistency across commands and enables proper validation of orchestration logic.

Agent orchestration in slash commands coordinates sub-agents by defining workflow steps, routing inputs between agents, and managing tool permissions. Structure each orchestration layer with explicit step sequencing and output handoffs.

Yes. Slash commands support workflow entry points for deployment, code review, and automation tasks. Define the command with parameter hints, tool permissions for each step, and orchestration logic to automate multi-stage processes end-to-end.

Enforce safety gates through tool permission restrictions, argument validation patterns, and standardized command structure. This prevents unintended agent actions and ensures commands execute within defined boundaries and approved tool access.

Refactor commands by optimizing argument parsing logic, reducing redundant orchestration steps, and streamlining tool permission declarations. Use the standardized .claude/commands structure to identify bottlenecks in workflow coordination.