skills.rest
Mosaic-agent
Mosaic-agentOFFICIAL
Explore all skills by this author →
GitHub

caveman

Compress model responses into caveman-style prose while preserving technical content.

36|7|Updated Feb 22, 2026
Search Tags:#style#llm#persistence#conversation#token-optimization#prompt-engineering#caveman
One-click install
npx skills add https://github.com/Mosaic-agent/Mosaic-fund-agent --skill caveman-mosaic-agent
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: caveman
Source: https://github.com/Mosaic-agent/Mosaic-fund-agent/tree/main/.agents/skills/caveman
Command: npx skills add https://github.com/Mosaic-agent/Mosaic-fund-agent --skill caveman-mosaic-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Chat sessions often include filler, hedging, and verbose replies that waste tokens and distract from the core content. This Skill compresses model responses into caveman-style prose while preserving exact technical content.

Core Features & Use Cases

  • Reduces token usage by removing filler language while keeping all technical details intact.
  • Supports six intensity levels (lite, full, ultra, wenyan-lite, wenyan-full, wenyan-ultra) and persists across a session.
  • Auto-applies in conversations demanding concise, precise responses or security-sensitive prompts.

Quick Start

Enable caveman mode by issuing '/caveman' to begin in full mode.

Frequently Asked Questions about caveman

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

FAQPage Schema
How do I reduce token usage in conversations without losing technical details?▼

Use caveman mode to compress LLM responses into terse prose, stripping filler language while preserving all exact technical details. This significantly lowers token consumption and removes distracting hedging during long chat sessions.

What is the best way to make LLM responses more concise during a chat session?▼

Activate caveman mode to enforce terse, exact responses across your entire conversation. It applies an auto-clarity rule to compress model replies while consistently retaining the core technical payload.

How do I enable and control conversation persistence for compressed responses?▼

Issue the '/caveman' command to start in full mode. The Skill uses per-response state persistence and frontmatter-driven configuration, ensuring the chosen compression setting persists until you explicitly change it.

Can I adjust the intensity of prompt engineering for token optimization?▼

Yes, token optimization intensity can be set across six levels: lite, full, ultra, wenyan-lite, wenyan-full, and wenyan-ultra. The default is full, and the chosen compression level persists throughout the ongoing conversation.

Does caveman-style response compression work automatically for security-sensitive prompts?▼

Yes, the compression auto-applies in conversations demanding concise, precise responses or handling security-sensitive prompts. It implements auto-clarity rules and per-response state persistence to ensure consistent behavior.

Related Skills

Explore More Content & Communication Skills→
AbdullahMalik17AbdullahMalik17
9

streaming-llm-responses

Manage streaming response lifecycle events and thread state in AI chat UIs.

Community
Advanced
tyevanstyevans
14

expand

Expand sparse findings into detailed descriptions with cited evidence.

Community
Intermediate
wangyouanwangyouan

context-compression

Compress large token contexts to minimize information loss in agent sessions.

Community
Advanced
Manvendra08Manvendra08
1

token-efficiency

Reduce token consumption and context overflow in LLM interactions.

Community
Basic