caveman

Compress AI responses by removing non-essential linguistic fluff.

1|Updated May 12, 2026
One-click install
npx skills add https://github.com/Manvendra08/TradingBot --skill caveman-manvendra08
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: caveman
Source: https://github.com/Manvendra08/TradingBot/tree/main/_agent/skills/caveman
Command: npx skills add https://github.com/Manvendra08/TradingBot --skill caveman-manvendra08

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill eliminates the problem of excessive token consumption in AI responses by removing all non-essential fluff, filler words, and pleasantries while retaining 100% of technical accuracy and critical information, making responses far more efficient for technical workflows.

Core Features & Use Cases

  • 6 Configurable Intensity Levels: Choose from lite, full, ultra, wenyan-lite, wenyan-full, and wenyan-ultra to match your compression needs, from professional tight phrasing to extreme classical Chinese compression.
  • Auto-Trigger Support: Automatically activates when you request token efficiency, say "caveman mode", or use the /caveman command, no manual setup required.
  • Safety Boundaries: Automatically disables caveman mode for security warnings, irreversible action confirmations, and multi-step sequences where fragment order could cause misread, then resumes after the critical section is complete.
  • Use Case: When debugging a production authentication bug, use caveman mode to get the root cause and fix in 1/4 the tokens of a standard response, with no loss of technical detail.

Quick Start

Request caveman mode activation when you need concise, technically accurate responses for debugging, code reviews, or quick technical queries to cut token usage by up to 75%.

Frequently Asked Questions about caveman

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

FAQPage Schema
How do I reduce AI token usage in technical responses without losing code accuracy?

Response compression eliminates linguistic fluff and filler while preserving technical terminology and code blocks, reducing AI token consumption by approximately 75% for debugging and technical Q&A workflows without sacrificing accuracy.

What is caveman mode for AI token reduction?

Caveman mode is a response compression technique that removes non-essential pleasantries and fluff from technical communication. It retains 100% of technical accuracy and critical information while cutting token usage by 75% across six configurable intensity levels.

How do I configure the intensity level for response compression?

Configure response compression by selecting from six intensity levels: lite, full, ultra, wenyan-lite, wenyan-full, and wenyan-ultra. These range from professional tight phrasing to extreme classical Chinese compression, matching your specific token efficiency needs.

Does response compression affect safety warnings in technical system status updates?

Response compression automatically disables itself for security warnings, irreversible action confirmations, and multi-step sequences to prevent misreads. It resumes normal compressed output only after the critical safety section is complete.

Can I auto-trigger token efficiency mode during code reviews and debugging?

Yes, you can auto-trigger token efficiency mode by explicitly requesting it, saying "caveman mode", or using the /caveman command. No manual setup is required to start receiving concise, technically accurate responses for code reviews and debugging.

When should I avoid using extreme classical Chinese compression for technical Q&A?

Avoid extreme classical Chinese compression (wenyan-ultra) for multi-step sequences where fragment order could cause misread, or when communicating irreversible action confirmations. Use lower intensity levels for these safety-critical technical Q&A scenarios.