What problem does it solve? The claude-flow v3 CLI suffers from monolithic files (index.ts at 108KB), basic command parsing, shallow hooks integration, and no intelligent workflow automation, making it slow and hard to maintain. ## Core Features & Use Cases - Modular Command Architecture: Decomposes large CLI files into focused command modules under 500 lines each, with a registry supporting categories, aliases, and fuzzy matching. - Interactive Prompt System: Provides context-aware select, multiselect, input, confirm, and progress-bar prompts using inquirer and cli-progress. - Enhanced Hooks & Learning Integration: Connects CLI lifecycle events (command_start, command_end, command_error) to trajectory tracking, reward calculation, and pattern storage via AgentDB and agentic-flow hooks. - Workflow Automation & Completion: Orchestrates multi-step workflows with dependencies, retry policies, and learning-based intelligent command auto-completion. - Use Case: A developer runs claude-flow swarm init --interactive and is guided through topology selection and agent configuration, while hooks record execution metrics and suggest optimizations based on similar past runs. ## Quick Start Ask the AI to analyze the current claude-flow CLI structure and implement modular commands, interactive prompts, and hooks integration using the cli-hooks-developer agent.