sublime

Coordinate disciplined code-craft workflows for AI coding agents.

Updated Apr 27, 2026
One-click install
npx skills add https://github.com/nokusukun/sublime --skill sublime
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sublime
Source: https://github.com/nokusukun/sublime/tree/main/sublime
Command: npx skills add https://github.com/nokusukun/sublime --skill sublime

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Provides a disciplined code-craft framework for AI coding agents, giving them a shared vocabulary, explicit guidelines, and an anti-pattern catalog to avoid the common slop that plagues automated coding prompts.

Core Features & Use Cases

  • Shared verb vocabulary: a concise set of coding disciplines (Create, Evaluate, Refine, Simplify, Harden) that align prompts with craft and outcome.
  • Comprehensive guidelines: extensive principles for naming, control flow, data modeling, errors, interfaces, dependencies, tests, and comments, plus a named anti-pattern catalog to detect and refuse slop.
  • Extensible language support: core skill is language-agnostic with language-specific extensions under languages/, enabling targeted discipline per language.
  • Context and governance: explicit context-gathering and coding-standards prompts to load project context and enforce consistent behavior across tasks.

Quick Start

Prompt Sublime with craft for plan-then-write, teach for project context, or extract to lift patterns.

Frequently Asked Questions about sublime

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

FAQPage Schema
How do I stop AI coding agents from generating poor quality code with anti-patterns?

To prevent AI coding agents from generating anti-patterns, apply a disciplined code-craft framework with a shared verb vocabulary and a named anti-pattern catalog. Sublime coordinates this by enforcing explicit guidelines for naming, control flow, and data modeling.

What is context-loading for AI coding and how does it improve code generation?

Context-loading for AI coding is the process of gathering project context to enforce consistent behavior. Sublime supports craft, teach, and extract modes to load project standards, ensuring prompts align with Create, Evaluate, Refine, Simplify, and Harden disciplines.

How do I enforce naming principles and coding standards in AI-generated code?

You can enforce naming principles and coding standards in AI-generated code by using explicit context-gathering prompts. Sublime provides comprehensive guidelines for naming, control flow, errors, and interfaces to align AI outputs with project standards.

Does this code-craft framework work for multiple programming languages or is it language-specific?

This code-craft framework is language-agnostic at its core but supports language-specific extensions. Sublime enables targeted discipline per language through its extensible languages directory, applying across writing, reviewing, refactoring, and auditing tasks.

What do I need to set up before using Sublime for AI-driven coding workflows?

Before using Sublime for AI-driven coding, you need a root SKILL.md file with YAML frontmatter containing name and description. Optional scripts, references, and assets directories can be added alongside a body explaining operational prompts and governance.

When should I avoid using a shared vocabulary framework for AI coding agents?

You should avoid using a shared vocabulary framework for AI coding agents during rapid prototyping or throwaway scripting where strict governance is unnecessary. Sublime is designed for disciplined code-craft across writing, reviewing, refactoring, and hardening production code.