caveman

Compress assistant responses into terse technical replies with configurable intensity levels.

4|2|Updated May 18, 2022

Search Tags:#prompt-engineering #token-efficiency #concise #caveman #technical-communication #wenyan #assistant-style

Ready to run

npx skills add https://github.com/pjmagee/starwars-data --skill caveman-pjmagee

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: caveman
Source: https://github.com/pjmagee/starwars-data/tree/main/.agents/skills/caveman
Command: npx skills add https://github.com/pjmagee/starwars-data --skill caveman-pjmagee

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reduce token consumption from conversational AI while preserving precise technical content so users can get concise, cost-efficient answers without losing correctness.

Core Features & Use Cases

Persistent terse mode that compresses language across responses using configurable intensity levels for tone and compression.
Rule-driven removal of articles, filler, pleasantries, and hedging while keeping technical terminology and code blocks intact, with explicit safety exceptions for destructive or ambiguous operations.
Auto-triggering on token-efficiency requests or keywords, and manual control via simple commands; ideal for developers, code reviewers, and engineers who need compact diagnostics, debugging hints, and succinct design explanations.

Quick Start

Say caveman mode full or send /caveman full to enable terse, technically accurate replies.

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 AI responses without losing technical accuracy?▼

To reduce token usage without losing technical accuracy, apply response compression that removes filler words and pleasantries while preserving code blocks and technical terminology. This ensures concise, cost-efficient answers for developer interactions and debugging guidance.

What is terse prompt engineering and how does it work for code explanations?▼

Terse prompt engineering compresses assistant language by rule-driving the removal of articles and hedging while keeping technical terminology intact. It works by applying configurable intensity levels to strip non-essential text from code explanations and debugging hints.

How do I enable concise mode for technical support interactions?▼

Enable concise mode for technical support interactions by sending a command like 'caveman mode full' or '/caveman full'. This triggers persistent session state that auto-compresses subsequent replies, ensuring compact diagnostics across the conversation.

Can I configure different intensity levels for token efficiency in developer documentation summaries?▼

Yes, you can configure token efficiency intensity levels for developer documentation summaries using lite, full, ultra, wenyan-lite, wenyan-full, and wenyan-ultra modes. These levels control the compression tone applied to the summarized technical content.

Are there safety limitations when compressing AI responses for debugging guidance?▼

When compressing AI responses for debugging guidance, safety limitations include explicit exceptions for destructive or ambiguous operations. The compression mechanism avoids stripping critical context that could compromise operational safety or introduce technical ambiguity.