caveman-compress

Compress natural language memory files into a compact caveman format.

Updated Feb 24, 2026
One-click install
npx skills add https://github.com/VXerys/marketiv-web --skill caveman-compress-vxerys
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: caveman-compress
Source: https://github.com/VXerys/marketiv-web/tree/main/.qwen/skills/caveman-compress
Command: npx skills add https://github.com/VXerys/marketiv-web --skill caveman-compress-vxerys

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill compresses project memory files, like CLAUDE.md, todos, and preferences, into a caveman format, reducing the number of tokens needed per session.

Core Features & Use Cases

  • Token Reduction: Compresses natural language files to reduce the amount of tokens used per session.
  • File Backup: Saves a human-readable backup of the original file as <filename>.original.md.
  • File Type Support: Works with .md, .txt, .rst, .typ, .typst, .tex files.
  • Code Block Preservation: Does not modify code blocks or inline code, ensuring that technical content remains intact.
  • Use Case: For developers working with Claude Code, this Skill can significantly reduce the token cost for each project session, leading to more efficient work and cost savings.

Quick Start

Compress the CLAUDE.md file using the caveman-compress skill.

Frequently Asked Questions about caveman-compress

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

FAQPage Schema
How do I reduce token usage for Claude Code memory files?

To reduce token usage for Claude Code memory files, you can compress natural language files like CLAUDE.md into a compact caveman format. This preserves file structure and technical content while lowering the token count processed per session.

Does memory file compression preserve code blocks and inline code?

Yes, memory file compression preserves code blocks and inline code without modification. The caveman format conversion targets natural language text, ensuring that all technical content and structural formatting remain fully intact.

What file types can I compress for token savings in Claude Code?

You can compress Markdown, plain text, reStructuredText, Typst, and LaTeX file types for token savings. Supported extensions include .md, .txt, .rst, .typ, and .tex files used as project memory files.

Do I need Python to compress project memory files into the caveman format?

Yes, you need Python 3.10 or higher installed locally to compress project memory files into the caveman format. The Skill relies on local execution scripts to process and convert the natural language text.

Can I restore the original content after compressing a CLAUDE.md file?

Yes, you can restore original content because the compression process saves a human-readable backup. The original file is automatically saved with an .original.md extension before the compressed version replaces it.

What is the best way to back up project memory files before token compression?

The best way to back up project memory files before token compression is to let the Skill handle it automatically. It saves a complete human-readable backup of the original file as <filename>.original.md during the conversion process.