Verified:GitHub

caveman

Compress AI responses into caveman-style prose with intensity levels.

Updated Aug 23, 2026

Search Tags:#debugging #token efficiency #response compression #prompt formatting #classical style #safety-aware verbosity

Ready to run

npx skills add https://github.com/junevm/avtoolz --skill caveman-junevm

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: caveman
Source: https://github.com/junevm/avtoolz/tree/main/.agents/skills/caveman
Command: npx skills add https://github.com/junevm/avtoolz --skill caveman-junevm

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents wasted output tokens by compressing responses into caveman-style prose while preserving technical accuracy and exact symbols.

Core Features & Use Cases

Session-persistent terse mode: Keeps the compression style active across the conversation until explicitly turned off.
Intensity levels: Supports lite, full, ultra, and classical Chinese variants (wenyan-*), letting you trade readability for brevity.
Safety-aware clarity: Automatically switches back to normal prose for security warnings, irreversible actions, ambiguous multi-step sequences, and repeated questions.

Quick Start

Ask for caveman mode by saying: /caveman ultra.

Frequently Asked Questions about caveman

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

FAQPage Schema

How do I reduce token usage in AI responses without losing technical details?▼

To reduce token usage without losing technical details, you can compress AI responses into caveman-style prose. This preserves code blocks, exact symbols, and debugging accuracy while significantly cutting output length.

Can I use response compression for multi-turn debugging and Q&A sessions?▼

Yes, response compression works for multi-turn debugging and Q&A sessions. The terse mode stays active across the entire conversation until you explicitly turn it off, maintaining technical accuracy throughout.

What intensity levels are available for token compression in prompt formatting?▼

Available intensity levels for token compression include lite, full, ultra, and classical Chinese variants (wenyan-*). These levels let you trade readability for brevity depending on your needs.

Does safety-aware verbosity automatically switch back to normal prose for security warnings?▼

Yes, safety-aware verbosity automatically switches back to normal prose for security warnings. It also enforces clarity for irreversible confirmations, ambiguous multi-step sequences, and repeated questions to prevent misunderstandings.

How do I activate caveman mode to start compressing AI output tokens?▼

To activate caveman mode and start compressing AI output tokens, simply type `/caveman ultra` or specify your desired intensity level. The compression style will remain active for the duration of your session.