craft-ace

Implement and review features locally with Gemma under Opus orchestration.

17|Updated Mar 30, 2026
One-click install
npx skills add https://github.com/alexiolan/craft-skills --skill craft-ace
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: craft-ace
Source: https://github.com/alexiolan/craft-skills/tree/main/skills/craft-ace
Command: npx skills add https://github.com/alexiolan/craft-skills --skill craft-ace

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Cost-optimized Craft ACE replaces cloud-only craft with a locally hosted Gemma-powered implementation and review pipeline, dramatically reducing compute costs while preserving quality.

Core Features & Use Cases

  • Local implementation and reviews powered by Gemma, delivering significant cost savings and faster iteration.
  • Opus orchestration with Sonnet fallback to handle integration and edge tasks, ensuring continuity when Gemma hits limits.
  • Profile-driven guidance that follows the canonical craft workflow (Brainstorm → Plan → Develop → Browser Test → Report) with cost-aware execution.

Quick Start

Describe your target feature and let the local Gemma-powered craft pipeline implement and review it.

Frequently Asked Questions about craft-ace

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I reduce LLM orchestration costs for local feature development?

You can reduce LLM orchestration costs by using a locally hosted Gemma model for coding and reviews, while Opus orchestrates the workflow and Sonnet handles fallback tasks to ensure continuity.

What is the craft pipeline workflow for locally hosted models?

The craft pipeline workflow is a profile-driven, end-to-end feature delivery sequence: Brainstorm, Plan, Develop, Browser Test, and Report, specifically optimized for cost-aware execution using local models.

Do I need LM Studio to run the Gemma-powered craft pipeline?

Yes, you need LM Studio availability to run the Gemma-powered craft pipeline, as it provides the local hosting environment required for implementation and review tasks.

Can I use Opus and Sonnet as fallbacks when local Gemma hits limits?

Yes, Opus orchestrates the pipeline and Sonnet provides fallback for edge tasks, ensuring continuity when the local Gemma model hits its operational limits during development.

What are the limitations of replacing cloud-based craft with local Gemma?

Limitations of using local Gemma include hitting model capacity limits during complex tasks, requiring Sonnet fallback for edge cases, and needing LM Studio availability for local hosting.

Best way to automate feature delivery with local LLMs and cloud fallback?

The best way is a profile-driven pipeline where local Gemma handles implementation and reviews, Opus orchestrates the process, and Sonnet provides cloud fallback for integration tasks.