Compress AI chat responses into caveman-style fragments to reduce token consumption.
npx skills add https://github.com/tDalile/dotfiles --skill caveman-tdalile
Please help me install this Agent Skill. Skill: caveman Source: https://github.com/tDalile/dotfiles/tree/main/agents/skills/caveman Command: npx skills add https://github.com/tDalile/dotfiles --skill caveman-tdalile
High token consumption inflates costs and slows down AI conversations; this Skill compresses replies to a bare‑bones caveman style, drastically reducing token count while preserving technical accuracy.
/caveman ultra to deliver concise, token‑efficient steps without redundant wording.Activate caveman mode with /caveman full to receive terse, token‑saving replies.
High-intent search queries and answers about installing and using this skill.
You can reduce token usage by applying compressed caveman-style responses that drop articles and filler words. This technique slashes token consumption by up to 75% while preserving technical accuracy through rule-based syntax compression.
Brevity compression supports lite, full, ultra, and classical Chinese (wenyan) intensity levels. Each mode balances detail and token efficiency differently, allowing you to select the exact compression strength needed for your conversation.
Persistent mode keeps caveman-style compression active across multiple turns until explicitly turned off. Once activated with a command like /caveman full, the terse reply style stays engaged automatically throughout the session.
An auto-clarity safeguard automatically switches back to normal responses for security warnings, irreversible actions, or clarification requests. This ensures compressed replies never obscure critical safety information during technical troubleshooting.
You should avoid terse replies when handling security warnings, irreversible actions, or when clarification is requested. The auto-clarity safeguard detects these contexts and automatically restores normal, detailed responses to prevent misunderstandings.
The best way to get token-efficient troubleshooting steps is enabling an ultra compression mode. This delivers concise instructions by enforcing fragment patterns and dropping redundant wording while maintaining the required technical accuracy.
gonzalezpazmonicaCompress AI context windows using a 5-phase framework to reduce token usage.
JNRCHAYANCompresses assistant responses into terse, technically accurate language.
jaysooReduce token usage in AI responses while preserving technical accuracy.
ElliotJLTEstimate token costs for CSV, JSON, and TXT files before loading.