Enforce concise, high-signal responses in coding and debugging conversations.
npx skills add https://github.com/vessux/dotfiles --skill caveman-vessux
Please help me install this Agent Skill. Skill: caveman Source: https://github.com/vessux/dotfiles/tree/main/umbel/skills/pocock/caveman Command: npx skills add https://github.com/vessux/dotfiles --skill caveman-vessux
Ultra-compressed communication mode. Cuts token usage ~75% by dropping filler, articles, and pleasantries while keeping full technical accuracy. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", or invokes /caveman.
Activate caveman mode by saying caveman mode or invoking /caveman to force terse, high-signal replies.
High-intent search queries and answers about installing and using this skill.
To reduce conversational token usage during AI-assisted coding, you can activate a terse response mode that enforces concise, high-signal replies, dropping filler and articles to cut token consumption by roughly 75%.
Terse prompt engineering for interactive debugging is a deterministic reply style transformation that omits hedging and pleasantries, ensuring quick, precise guidance while maintaining technical accuracy across multi-turn planning contexts.
To trigger a concise reply style for rapid-fire development chats, use user-activated commands like saying "caveman mode", "less tokens", or invoking /caveman to force minimalist, high-signal outputs.
Minimalist AI response mode works with multi-turn planning conversations by safely handling multi-turn contexts to provide deterministic, accurate, and high-signal guidance throughout extended interactive sessions.
Ultra-compressed communication mode for AI assistants should not be used when detailed explanations, conversational pleasantries, or expanded hedging are required, as it strictly enforces minimalist, high-signal replies.
pjmageeDisplay a one-shot reference card for caveman conversational modes and commands.
dragoscirjanCompresses assistant responses into terse, high-signal language while preserving technical accuracy.
nathanvaleCreate and optimize hierarchical CLAUDE.md files for token efficiency.
ek33450505Compress assistant prose into concise output while preserving code blocks and values.
