Reduce AI conversation token usage while preserving technical accuracy and code blocks.
npx skills add https://github.com/nikbrunner/dots --skill caveman-nikbrunner
Please help me install this Agent Skill. Skill: caveman Source: https://github.com/nikbrunner/dots/tree/main/common/.agents/skills/caveman Command: npx skills add https://github.com/nikbrunner/dots --skill caveman-nikbrunner
Ultra-compressed communication mode that cuts token usage while preserving full technical accuracy. It reduces verbosity in AI chats without sacrificing correctness, enabling faster, clearer technical exchanges.
Activate caveman mode by issuing /caveman and respond tersely with precise technical content.
High-intent search queries and answers about installing and using this skill.
You activate ultra-compressed mode by issuing the /caveman command or using trigger phrases like "caveman mode", which auto-activates terse, technically accurate responses.
You can reduce token usage without losing technical accuracy by applying an ultra-compressed communication mode that removes verbosity while preserving critical terms and code blocks.
The best way to get concise AI responses is to trigger a brevity mode using commands like /caveman, which forces the assistant to reply tersely with precise technical content.
Yes, you can adjust the brevity level using intensity levels like lite, full, and ultra to control the exact tone and compression of AI responses during technical discussions.
Compressed AI communication works safely across multiple turns by maintaining persistent operation with explicit deactivation and conservative phrasing to prevent technical misinterpretation.
You activate ultra-compressed mode by issuing the /caveman command or using trigger phrases like "caveman mode", which auto-activates terse, technically accurate responses.
se02035Compress communication to reduce token usage while maintaining technical accuracy.
yanochka11Record successful technical fixes as reusable debugging recipes.
PrepLabsAIAssess AI prompt engineering proficiency through multi-phase technical interviews.
tkerseyParse Codex JSONL session logs to extract skill usage and token counts.
