What are .claude/rules/?

These are modular instruction files that target specific directories or languages (e.g., python.md). They allow Claude to load relevant context automatically based on the file being edited, keeping the main memory lean.

What is a CLAUDE.md file?

A CLAUDE.md file is a project-specific markdown document that provides Claude with essential context, such as build commands, testing procedures, and coding style, to ensure consistent and accurate assistance.

How does this skill optimize token usage?

The skill applies compression techniques like converting verbose prose into tables and bullets, removing redundant information, and moving path-specific rules into modular files that only load when needed.

Can I store API keys in CLAUDE.md?

No. This skill includes a mandatory validation step that forbids sensitive information. Credentials and tokens should be handled via environment variables or secret managers, not persistent memory files.

Claude Project Memory Manager

Name: Claude Project Memory Manager
Author: digital-stoic-org

bydigital-stoic-org

Developer Tools

Optimizes and manages CLAUDE.md files and project rules to provide high-signal context for AI-driven development.

About

This skill streamlines the creation, updating, and optimization of project-specific memory files following Anthropic's best practices for Claude Code. It helps developers maintain a lean and efficient context hierarchy by managing the main CLAUDE.md file for universal project standards, modular .claude/rules/ for path-specific guidance, and local overrides for personal preferences. By enforcing a structured 'WHAT/WHY/HOW' organization and applying token-compression techniques, the skill ensures that Claude always has access to critical coding standards, build commands, and architectural patterns without exceeding instruction budgets or leaking sensitive credentials.

Key Features

Support for progressive disclosure using @imports and .claude/rules/ hierarchy
Modular rule management for path-specific and language-specific instructions
Token-efficiency optimization to keep project context under 200-500 tokens
Automated CLAUDE.md creation with structured WHAT/WHY/HOW templates
Mandatory security validation to prevent the inclusion of credentials or tokens
0 GitHub stars

Use Cases

Standardizing build, test, and deployment commands across a team development environment
Refining project memory to improve Claude's adherence to specific architectural patterns
Migrating monolithic instruction files into modular, path-specific rules to save tokens

About

Key Features

Support for progressive disclosure using @imports and .claude/rules/ hierarchy
Modular rule management for path-specific and language-specific instructions
Token-efficiency optimization to keep project context under 200-500 tokens
Automated CLAUDE.md creation with structured WHAT/WHY/HOW templates
Mandatory security validation to prevent the inclusion of credentials or tokens
0 GitHub stars

Use Cases

Standardizing build, test, and deployment commands across a team development environment
Refining project memory to improve Claude's adherence to specific architectural patterns
Migrating monolithic instruction files into modular, path-specific rules to save tokens