Skill Forge

If you're building Claude Code skills, you'll want this. It walks you through the entire process with a nine-step checklist: clarifying requirements, designing architecture, writing descriptions that actually trigger, building workflows with confirmation gates, and packaging. The real value is in the progressive loading strategy and the "Iron Law" concept, which forces you to identify the one mistake Claude will most likely make and guard against it upfront. Includes scripts for initialization and packaging, plus reference files on keyword bombing, anti-patterns, and parameter systems. Opinionated about token cost, which is good because most skills are bloated.

skills.sh installs: 1.2K
GitHub stars: 3.8K
last updated: Jul 14, 2026

View on GitHub

Install

$ npx -y skills add sanyuan0704/code-review-expert --skill skill-forge --agent claude-code

Installs into .claude/skills of the current project.

Files

Skill Forge

IRON LAW: Every line in a skill must justify its token cost. If it doesn't make the model's output better, more consistent, or more reliable — cut it.

What is a Skill

A skill is an "onboarding guide" for Claude — transforming it from a general-purpose agent into a specialized one with procedural knowledge, domain expertise, and bundled tools.

skill-name/
├── SKILL.md           # Required: workflow + instructions (<500 lines)
├── scripts/           # Optional: deterministic, repeatable operations
├── references/        # Optional: loaded into context on demand
└── assets/            # Optional: used in output, never loaded into context

Default assumption: Claude is already very smart. Only add what Claude doesn't already know. Challenge every paragraph: "Does this justify its token cost?"