craft-local

Orchestrate end-to-end craft workflows with local LLM reviews via LM Studio.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables end-to-end software feature development using local LLM reviews to improve spec quality, plan accuracy, and post-develop validation, while keeping processing local to your environment.

Core Features & Use Cases

  • End-to-end Craft pipeline with Brainstorm, Plan, Develop, Browser Test, and Report stages.
  • Local-LM reviews via LM Studio gating ensure spec, plan, and post-develop checks.
  • Wrapper-based local workflow preserves pre-1.1 Craft behavior while enabling on-prem governance.

Quick Start

Describe your feature using /craft to start the local-LM assisted pipeline.

Frequently Asked Questions about craft-local

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

FAQPage Schema
How do I automate feature development with a local LLM?

You can automate feature development with a local LLM by running the craft-local profile to orchestrate brainstorming, planning, development, browser testing, and reporting stages end-to-end.

What is local-review gating in software development workflows?

Local-review gating is a mechanism that uses local LLMs via LM Studio to enforce spec, plan, and post-development checks, ensuring deterministic task orchestration and stepwise execution.

Does the craft-local workflow require LM Studio integration?

Yes, the craft-local workflow requires LM Studio integration to run local LLM reviews and gate feature development stages within your on-premises environment.

Can I run browser testing as part of an end-to-end local LLM pipeline?

Yes, browser testing is included as a dedicated stage in the end-to-end local LLM pipeline, running automatically after the development phase to validate post-develop outputs.

What is the best way to keep LLM processing local during software planning?

To keep LLM processing local during software planning, use a wrapper-based local workflow that preserves deterministic task orchestration while applying local LLM reviews to gate spec and plan accuracy.

Why does local-review gating require access to core craft definitions?

Local-review gating requires access to core craft definitions in skills/craft because the craft-local profile relies on these foundational definitions to execute deterministic task orchestration and stepwise validation.