What is the 'Extract to skill' action?

If a section of your CLAUDE.md contains a reusable workflow or script that could benefit other projects, this skill suggests moving it into a standalone Claude Code skill.

What is progressive disclosure for CLAUDE.md?

It is a design principle where you keep essential instructions and core rules in the primary CLAUDE.md file while moving detailed examples, troubleshooting guides, and procedures to external reference files.

How does reducing CLAUDE.md size help?

Smaller configuration files reduce token usage and prevent the AI from being 'distracted' by irrelevant details, leading to faster response times and more accurate task execution.

Does this skill delete my project documentation?

No. The skill moves detailed content to the ~/.claude/references/ directory and leaves a concise pointer in the original CLAUDE.md so the AI still knows where to find the information when required.

Can I choose which sections to keep or move?

Yes. The skill generates an Optimization Proposal first, allowing you to review and approve which sections are kept, moved, or extracted before any changes are applied.

CLAUDE.md Context Optimizer

Name: CLAUDE.md Context Optimizer
Author: daymade

bydaymade

•

147

•

Productivity & Workflow

Streamlines CLAUDE.md files by applying progressive disclosure principles to reduce context overhead and improve AI response efficiency.

The CLAUDE.md Context Optimizer is designed to manage and reduce the size of your project's configuration files without losing critical information. By auditing section lengths and complexity, it identifies content that is better suited for external reference files or standalone skills. This progressive disclosure approach ensures that Claude's immediate context remains focused on core principles and frequently used commands, while detailed procedures, infrastructure logs, and verbose code examples are moved to a structured references directory, significantly improving token efficiency and model performance.

Key Features

01Seamless pointer insertion to maintain AI access to externalized content

02Intelligent classification of content into 'Keep', 'Reference', 'Skill', or 'Remove'

03Standardized extraction of verbose documentation to ~/.claude/references/

04Automated audit of CLAUDE.md sections with line-count analysis

05Optimization proposals with projected line-count reduction percentages

06147 GitHub stars

Use Cases

01Identifying and extracting reusable development patterns into global Claude Code skills

02Standardizing documentation by separating high-level rules from low-level implementation details

03Reducing context bloat in projects where CLAUDE.md has grown too large to be token-efficient

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add daymade/claude-code-skills claude-md-progressive-disclosurer

For use in Claude.ai and ChatGPT

Key Features

01Seamless pointer insertion to maintain AI access to externalized content

02Intelligent classification of content into 'Keep', 'Reference', 'Skill', or 'Remove'

03Standardized extraction of verbose documentation to ~/.claude/references/

04Automated audit of CLAUDE.md sections with line-count analysis

05Optimization proposals with projected line-count reduction percentages

06147 GitHub stars

Use Cases

01Identifying and extracting reusable development patterns into global Claude Code skills

02Standardizing documentation by separating high-level rules from low-level implementation details

03Reducing context bloat in projects where CLAUDE.md has grown too large to be token-efficient

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add daymade/claude-code-skills claude-md-progressive-disclosurer

For use in Claude.ai and ChatGPT