Graph Workflow Orchestrator

概要

Graph Workflow Orchestrator is a lightweight framework designed to transform complex Claude Code tasks into structured, visualizable workflows. By leveraging graph abstraction and intelligent model routing, it automatically assigns subtasks to the most efficient Claude model—Haiku for exploration, Sonnet for analysis, or Opus for critical reasoning. This approach results in up to 75% cost savings and 70% less context usage, making it an essential tool for developers building sophisticated RAG pipelines, multi-agent systems, or deep repository analysis tools.

主な機能

• Graph-based workflow abstraction for visual task dependency management
• Intelligent context-efficiency reducing token usage by up to 70%
• 0 GitHub stars
• Automatic multi-model routing (Haiku, Sonnet, Opus) based on task complexity
• Native Claude Code integration with zero external framework dependencies
• Parallel task execution through automatic topological sorting of independent nodes

ユースケース

• Optimizing expensive automated tasks by offloading discovery phases to Claude 3 Haiku
• Architecting complex RAG pipelines for large-scale codebase exploration
• Creating multi-agent 'Perceive-Reason-Act' workflows with clear separation of concerns