caveman

Strip filler and hedging from AI responses while preserving technical accuracy.

15|5|Updated May 17, 2026
One-click install
npx skills add https://github.com/cropsgg/skills --skill caveman-cropsgg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: caveman
Source: https://github.com/cropsgg/skills/tree/main/skills/productivity/caveman
Command: npx skills add https://github.com/cropsgg/skills --skill caveman-cropsgg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ultra-compressed communication mode that drops filler, hedging, and meta-commentary while preserving full technical accuracy. Reduces token usage ~75% in tight-context scenarios, enabling rapid iteration without sacrificing precision.

Core Features & Use Cases

  • Strip filler, hedging, and meta-commentary from outputs to produce crisp, signal-focused results.
  • Enforce imperative voice and concise structure (headlines, bullet lists, and explicit file/path references).
  • Use in handoffs, runbooks, technical specs, or any context with restricted tokens where raw signal is required.

Quick Start

Enable caveman mode and respond with ultra-compact, technically precise output for the given prompt.

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 LLM outputs while preserving technical accuracy?

Reduce token usage by stripping filler, hedging, and meta-commentary from LLM outputs. This rule-based compression enforces imperative voice and concise formatting, cutting token consumption by roughly 75% in tight-context workflows without sacrificing precision.

What is the best way to format technical runbooks for rapid iteration cycles?

Format technical runbooks for rapid iteration by enforcing imperative voice, concise structure, and explicit file/path references. Stripping meta-commentary and filler ensures raw signal delivery, which is critical when operating under strict token limits.

Can I use ultra-compact prompt design for tight-context technical specs?

Yes, ultra-compact prompt design works for tight-context technical specs. It applies rule-based stripping of filler and hedging to deliver crisp, signal-focused results, enabling rapid iteration and handoffs without exceeding token limits.

Does stripping meta-commentary from AI responses affect path-exact references?

Stripping meta-commentary does not affect path-exact references. The compression process preserves full technical accuracy by enforcing imperative voice and explicit file/path references while removing only filler and hedging from the output.

When should I avoid using conciseness-focused prompt compression?

Avoid conciseness-focused prompt compression when outputs require conversational tone, detailed explanations, or extensive hedging. The tool is designed for raw signal delivery in technical specs and runbooks, stripping meta-commentary entirely.