Claude Code Memory & Instructions System

Manage project context with hierarchical CLAUDE.md files and scoped rules.

18|2|Updated Dec 12, 2025
One-click install
npx skills add https://github.com/markus41/claude --skill claude-code-memory-instructions-system
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Claude Code Memory & Instructions System
Source: https://github.com/markus41/claude/tree/main/plugins/claude-code-expert/skills/memory-instructions
Command: npx skills add https://github.com/markus41/claude --skill claude-code-memory-instructions-system

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides a comprehensive system for managing project instructions, rules, and memory, ensuring the AI consistently understands and adheres to your project's specific requirements and learned patterns.

Core Features & Use Cases

  • Centralized Instructions: Define project build commands, code conventions, and architecture in CLAUDE.md.
  • Enforceable Rules: Implement scoped rules for specific file types or global project guidelines using .claude/rules/.
  • Persistent Memory: Automatically save and load key project discoveries, patterns, and user preferences to improve AI performance over time.
  • Self-Healing: The lessons-learned system automatically captures errors, suggests fixes, and prevents recurrence.
  • Use Case: Ensure all new TypeScript files adhere to strict typing and use functional components by defining these as path-scoped rules, while global rules enforce conventional commits for all Git operations.

Quick Start

Use the Claude Code system to add a new rule for all markdown files to enforce a maximum line length of 100 characters.

Frequently Asked Questions about Claude Code Memory & Instructions System

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

FAQPage Schema
How do I persist AI context and project rules across different development sessions?

You can persist AI context across development sessions by using hierarchical CLAUDE.md files to store build commands and architectural decisions, alongside an auto-memory system that automatically saves and loads key project discoveries.

How do I enforce specific coding conventions for different file types in my project?

You can enforce coding conventions for different file types by creating scoped rules within the .claude/rules/ directory, allowing you to apply specific guidelines like strict typing for TypeScript files or line length limits for markdown files.

How do I prevent recurring AI coding errors and build failures?

To prevent recurring AI coding errors, you can implement a self-healing loop via lessons-learned documentation that automatically captures past errors, suggests fixes, and ensures the AI does not repeat the same mistakes.

What is the best way to manage global project instructions versus path-specific rules?

The best way to manage project instructions is to define global guidelines and build commands in a centralized CLAUDE.md file, while using the .claude/rules/ directory to enforce path-specific rules for individual file types or directories.

Can I use CLAUDE.md files to maintain consistent architectural decisions across a large project?

Yes, you can use a hierarchical structure of CLAUDE.md files to maintain consistent architectural decisions, ensuring the AI understands and adheres to your project's specific requirements and learned patterns across all directories.

Do I need any external dependencies to set up an auto-memory system for my AI coding assistant?

No external dependencies are required to set up the auto-memory system, as it functions by automatically saving and loading user preferences and project discoveries directly through the CLAUDE.md file structure.