skills.rest
Sambhav242005
Sambhav242005COMMUNITY
Explore all skills by this author →
WebsiteGitHub

caveman

Compresses AI responses into terse caveman-style prose while preserving technical accuracy.

Updated Aug 7, 2026
Search Tags:#concise output#llm efficiency#token compression#classical chinese#response style#caveman mode#prompt instruction
One-click install
npx skills add https://github.com/Sambhav242005/Major-Project --skill caveman-sambhav242005
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: caveman
Source: https://github.com/Sambhav242005/Major-Project/tree/main/.continue/skills/caveman
Command: npx skills add https://github.com/Sambhav242005/Major-Project --skill caveman-sambhav242005

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Long, filler-heavy AI responses waste output tokens and slow down reading. This Skill compresses every model response into terse caveman-style prose, cutting roughly 65% of output tokens while keeping all technical details, code, and error strings exact. ## Core Features & Use Cases - Six intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, and wenyan-ultra, including classical Chinese compression modes. - Session persistence: the mode stays active across all turns until explicitly stopped with "stop caveman" or "normal mode". - Auto-clarity rule: automatically reverts to normal prose for security warnings, irreversible-action confirmations, and ambiguous multi-step sequences, then resumes compression. - Use Case: During a long debugging session, enable full mode so every explanation arrives as short fragments with exact code and error strings, reducing token usage without losing technical substance. ## Quick Start Ask the AI to enable caveman mode by saying "use caveman mode" or invoking /caveman, then continue the conversation normally.

Frequently Asked Questions about caveman

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

FAQPage Schema
How do I reduce AI output tokens in chat responses?▼

Enable caveman mode by saying "use caveman" or invoking /caveman. It drops articles, filler, and hedging from every response, cutting roughly 65% of output tokens while keeping code, error strings, and technical terms exact.

How do I turn off caveman mode?▼

Say "stop caveman" or "normal mode" to revert to standard prose. The mode otherwise persists for the entire session and does not drift back on its own.

Does caveman mode change code blocks or error messages?▼

No. Code blocks, function names, API names, numbers, and exact error strings are never altered. Only prose is compressed; technical content stays verbatim.

What are the caveman intensity levels?▼

Six levels exist: lite (full sentences, no filler), full (default, drops articles), ultra (bare fragments), and wenyan-lite, wenyan-full, wenyan-ultra for classical Chinese compression up to 80-90% character reduction.

When does caveman mode stop compressing automatically?▼

It drops to normal prose for security warnings, irreversible-action confirmations, ambiguous multi-step sequences, and when the user repeats a question. Compression resumes after the clear part is done.

Does caveman mode affect written files like commits and docs?▼

No. Content persisted outside the chat, such as code, comments, commits, docs, and issue text, is written in normal prose. Compression applies only to conversational responses.

Related Skills

Explore More Software Engineering Skills→
viniciuscs84viniciuscs84

caveman

Compresses model responses into terse caveman-style prose while preserving technical accuracy.

Community
Basic
tiancaiamaotiancaiamao
5

session-logs

Search archived session JSONL logs to recover prior conversation context.

Community
Intermediate
belokonmbelokonm
8

performance-monitor

Monitor token usage, latency, and eval-driven quality in AI agent workflows.

Community
Advanced
Fidasek009Fidasek009
1

caveman

Compresses AI response output into terse fragments across six intensity levels while preserving technical accuracy.

Community
Intermediate