What problem does it solve?
Massive duplication and parallel implementations across claude-flow and agentic-flow@alpha make maintenance slow and inconsistent while blocking performance and feature parity.
Core Features & Use Cases
- Deep adapter-layer integration: Replaces parallel system components (coordinators, managers, schedulers) with a specialized extension of agentic-flow@alpha to sharply reduce code size.
- Feature-parity integration targets: Integrates SONA learning modes, Flash Attention performance improvements, and AgentDB cross-agent memory/search coordination.
- MCP tools and hooks wiring: Enables Claude/MCP tooling by leveraging available pre-built MCP tools and configuring multiple hook types for consistent task behavior.
- Migration with backward compatibility: Supports phased migration (adapter, system migration, cleanup) including dual operation and gradual feature migration to keep v2 functionality intact.
Quick Start
Ask the AI to produce the v3 adapter-layer and phased migration plan that reduces claude-flow duplication by replacing SwarmCoordinator, AgentManager, and TaskScheduler with agentic-flow@alpha equivalents.