local-cloud-hybrid

What problem does it solve?

Running all AI tasks through paid Claude API tiers leads to unnecessary high costs, especially for simple high-volume tasks like code formatting, batch labeling, and test stub generation that can be executed for free on a local model.

Core Features & Use Cases

• Intelligent Task Routing Decision Tree: Quickly determine whether to use a local LM Studio model or a Claude API tier (Haiku, Sonnet, Opus) based on task complexity, tool requirements, and cost efficiency.
• Windows 11 Local Setup Guide: Step-by-step instructions to install LM Studio, download recommended lightweight coding and general-purpose models, and configure the OpenAI-compatible local endpoint for use with Claude Code.
• Dual-Terminal Orchestration Pattern: Coordinate a Claude Sonnet orchestrator (for planning, delegation, and multi-step reasoning) with local or Haiku workers for isolated task execution, minimizing context bloat and cost.
• Cost Optimization Framework: Clear cost comparisons between local and Claude models, plus a savings strategy to batch simple tasks for local overnight execution.
• Zero-Token File Handoff Pattern: Pass task specifications and results between orchestrator and workers via disk files, eliminating extra token overhead for inter-agent communication.
• Use Case Example: Run 10,000 test stubs locally for $0 instead of ~$15 on Sonnet, while using Claude Opus for security reviews and Sonnet for multi-file code review.

Quick Start

Use this skill to configure a hybrid local LM Studio and Claude API setup on your Windows 11 machine that automatically routes simple high-volume tasks to a free local model and complex reasoning or orchestration tasks to the appropriate Claude tier to minimize API costs.