caveman-compress

Compress natural language text files to reduce Claude API input tokens.

Updated May 10, 2026

Search Tags:#file compression #token saving #token compression #text compression #CLAUE.md

Ready to run

npx skills add https://github.com/AhmadGehad/lfms --skill caveman-compress-ahmadgehad

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: caveman-compress
Source: https://github.com/AhmadGehad/lfms/tree/main/.kilocode/skills/caveman-compress
Command: npx skills add https://github.com/AhmadGehad/lfms --skill caveman-compress-ahmadgehad

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill compresses natural language files, such as CLAUDE.md and todos, into a format that uses fewer tokens, reducing the cost of token usage per session.

Core Features & Use Cases

Compression of Natural Language Files: Compresses files like CLAUDE.md, todos, and preferences to reduce the number of tokens needed.
Preservation of Structure: Keeps headings, code blocks, URLs, file paths, and technical terms intact.
Backup of Original Files: Saves a human-readable backup of the original file with the .original.md extension.
Use Case: For users who want to reduce the cost of using Claude's token-based API by compressing their project memory files.

Quick Start

Compress the CLAUDE.md file using the command: /caveman-compress CLAUDE.md

Frequently Asked Questions about caveman-compress

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

FAQPage Schema

How do I compress CLAUDE.md files to save Claude API tokens?▼

You can compress CLAUDE.md files to save Claude API tokens by running the `/caveman-compress CLAUDE.md` command. This reduces natural language input tokens while preserving headings, code blocks, URLs, and file paths.

Does text compression for Claude preserve technical terms and file paths?▼

Text compression for Claude preserves technical terms, file paths, URLs, headings, and code blocks. The Skill compresses natural language project memory files while maintaining the structural elements needed for accurate context.

Do I need Python to compress markdown files for token reduction?▼

Yes, you need Python installed to compress markdown files for token reduction. The Skill requires a Python environment and access to Claude's API to process and compress natural language text files.

What is the best way to reduce input tokens for Claude's API?▼

The best way to reduce input tokens for Claude's API is to compress natural language files like CLAUDE.md and todos into a compact format. This approach maintains technical structures while lowering token usage costs per session.

Will compressing project memory files delete my original markdown content?▼

Compressing project memory files will not delete your original markdown content. The Skill automatically saves a human-readable backup of the original file with the `.original.md` extension before applying token compression.