creating-claude-rules
About
This Claude Skill helps developers create and fix `.claude/rules/` files by ensuring correct frontmatter syntax, particularly using `paths` instead of `globs`. It provides guidance on proper glob patterns and avoids Cursor-specific fields like `alwaysApply`. Use this skill when setting up project-specific rules or migrating from Cursor to Claude's format.
Quick Install
Claude Code
Recommended/plugin add https://github.com/majiayu000/claude-skill-registrygit clone https://github.com/majiayu000/claude-skill-registry.git ~/.claude/skills/creating-claude-rulesCopy and paste this command in Claude Code to install this skill
Documentation
Creating Claude Rules
Overview
Rules in .claude/rules/ are modular instructions scoped to specific files via glob patterns. They load automatically with same priority as CLAUDE.md.
When to Use
- Creating new rules in
.claude/rules/ - Fixing rules that use wrong frontmatter (
globsinstead ofpaths) - Migrating Cursor rules to Claude format
- Organizing project-specific conventions
Quick Reference
| Field | Claude | Cursor |
|---|---|---|
| Path patterns | paths | globs |
| Always apply | Omit paths | alwaysApply: true |
| Description | Not documented | description |
Frontmatter
Path-Scoped Rules
---
paths:
- "src/api/**/*.ts"
- "tests/**/*.test.ts"
---
Or single pattern:
---
paths: src/**/*.{ts,tsx}
---
Global Rules
Omit frontmatter entirely - applies to all files:
# TypeScript Conventions
Use .js extensions in imports.
Common Mistakes
# ❌ WRONG - globs is Cursor format
---
globs:
- "**/*.ts"
---
# ✅ CORRECT - Claude uses paths
---
paths:
- "**/*.ts"
---
# ❌ WRONG - alwaysApply is Cursor-only
---
alwaysApply: true
---
# ✅ CORRECT - just omit paths for global rules
# (no frontmatter needed)
# ❌ WRONG - unquoted patterns
---
paths:
- **/*.ts
---
# ✅ CORRECT - quote glob patterns
---
paths:
- "**/*.ts"
---
Directory Structure
.claude/rules/
├── testing.md # Path-scoped or global
├── typescript.md
└── frontend/ # Subdirectories supported
└── react.md
Files discovered recursively. Use subdirectories to organize.
Glob Patterns
| Pattern | Matches |
|---|---|
**/*.ts | All .ts files anywhere |
src/**/* | Everything under src/ |
*.md | Markdown in root only |
**/*.{ts,tsx} | .ts and .tsx files |
{src,lib}/**/*.ts | .ts in src/ or lib/ |
Reference
https://code.claude.com/docs/en/memory#modular-rules-with-claude/rules/
GitHub Repository
Related Skills
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
evaluating-llms-harness
TestingThis Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.
langchain
MetaLangChain is a framework for building LLM applications using agents, chains, and RAG pipelines. It supports multiple LLM providers, offers 500+ integrations, and includes features like tool calling and memory management. Use it for rapid prototyping and deploying production systems like chatbots, autonomous agents, and question-answering services.
llamaguard
OtherLlamaGuard is Meta's 7-8B parameter model for moderating LLM inputs and outputs across six safety categories like violence and hate speech. It offers 94-95% accuracy and can be deployed using vLLM, Hugging Face, or Amazon SageMaker. Use this skill to easily integrate content filtering and safety guardrails into your AI applications.