How do I install the Create Rules skill?

Run npx skillmds@latest add diegosouzapw/create-rules in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Create Rules safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. Capability flags: reads secrets. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Create Rules?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Create Rules free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Create Rules?

diegosouzapw (@diegosouzapw) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Create Rules by diegosouzapw — Agent Skill

Name: Create Rules
Rating: 4.2 (5 reviews)
Author: diegosouzapw

Create Rules

Create or update Cursor Rules (.mdc files) and Skills (SKILL.md). Use when creating rules, adding coding standards, setting up conventions, updating .cursor/rules/, or converting rules to skills. Defines standard format, naming, frontmatter, token budget.

diegosouzapw Updated Sep 8, 2026 54 repo stars

File contents

Create Rules and Skills

Rule Types (3-Tier System)

Type	Frontmatter	When loaded	Filename suffix
Always	`alwaysApply: true`, description/globs blank	Every conversation	`-always.mdc`
Auto	`alwaysApply: false`, globs filled	When matching file is open	`-auto.mdc`
Agent	`alwaysApply: false`, description filled	Agent decides	`-agent.mdc`
Manual	`alwaysApply: false`, both blank	Only via @ mention	`-manual.mdc`

Token Budget (CRITICAL)

Total always-apply: TARGET < 5,000 tokens
alwaysApply: true ONLY for rules needed in EVERY chat
File-specific rules -> auto (globs)
Task-specific rules -> agent (description) or skill
Rule of Three: codify a rule ONLY after 3 repetitions of the same mistake

Naming Convention

{CATEGORY}-rule-name-{always|auto|agent|manual}.mdc

Categories: core, _base, protocol, standard, workflows, error

Rule Structure

See template in rule-template.mdc.

Description Format (Agent Rules)

ACTION-TRIGGER-OUTCOME format: "Применять при [TRIGGER]. [ACTION] для [OUTCOME]."

Universality Requirement

Rules must work in ANY project:

NO project entities -> use placeholders
NO project URLs/creds -> .cursor/.secrets/
NO project architecture -> AGENTS.md
YES: universal principles, abstract patterns

When to Create a Skill Instead of a Rule

Workflow with multiple steps -> Skill
Needs executable scripts -> Skill
Long procedural content (>200 lines) -> Skill
Only relevant for specific tasks -> Skill
Short constraint needed broadly -> Rule

After Creating/Updating

Re-read the file (self-check)
Output: Rule/Skill path, Type, Description

---name: create-rulesdescription: Create or update Cursor Rules (.mdc files) and Skills (SKILL.md). Use when creating rules, adding coding standards, setting up conventions, updating .cursor/rules/, or converting rules to skills. Defines standard format, naming, frontmatter, token budget.---# Create Rules and Skills## Rule Types (3-Tier System)| Type | Frontmatter | When loaded | Filename suffix ||------|-------------|-------------|-----------------|| **Always** | `alwaysApply: true`, description/globs blank | Every conversation | `-always.mdc` || **Auto** | `alwaysApply: false`, globs filled | When matching file is open | `-auto.mdc` || **Agent** | `alwaysApply: false`, description filled | Agent decides | `-agent.mdc` || **Manual** | `alwaysApply: false`, both blank | Only via @ mention | `-manual.mdc` |## Token Budget (CRITICAL)- Total always-apply: TARGET < 5,000 tokens- alwaysApply: true ONLY for rules needed in EVERY chat- File-specific rules -> auto (globs)- Task-specific rules -> agent (description) or skill- **Rule of Three**: codify a rule ONLY after 3 repetitions of the same mistake## Naming Convention`{CATEGORY}-rule-name-{always|auto|agent|manual}.mdc`Categories: core, _base, protocol, standard, workflows, error## Rule StructureSee template in [rule-template.mdc](assets/rule-template.mdc).## Description Format (Agent Rules)ACTION-TRIGGER-OUTCOME format:"Применять при [TRIGGER]. [ACTION] для [OUTCOME]."## Universality RequirementRules must work in ANY project:- NO project entities -> use placeholders- NO project URLs/creds -> .cursor/.secrets/- NO project architecture -> AGENTS.md- YES: universal principles, abstract patterns## When to Create a Skill Instead of a Rule- Workflow with multiple steps -> Skill- Needs executable scripts -> Skill- Long procedural content (>200 lines) -> Skill- Only relevant for specific tasks -> Skill- Short constraint needed broadly -> Rule## After Creating/Updating1. Re-read the file (self-check)2. Output: Rule/Skill path, Type, Description

Frequently asked questions