JuliusBrussee/caveman
Caveman
Caveman is a set of tools and configurations designed for large language model token optimization. It focuses on reducing the amount of data processed during AI interactions to lower costs and maximize the available context window.
The project implements a fragmented communication style that replaces full grammatical sentences with concise technical keywords. This approach extends to AI context optimization by condensing memory files and tool descriptions, and includes a specialized configuration for generating terse, one-line code reviews and short conventional commit messages.
The system includes monitoring tools to track real-time session token consumption and translate those figures into monetary cost savings based on session logs.
Features
- Token Optimization Utilities - Provides a comprehensive suite of tools to reduce token consumption in LLM prompts and responses.
- Output Compression - Implements a fragmented communication style to reduce the number of tokens generated by the AI.
- Context Compression - Condenses project documentation and memory files to fit more relevant information into the AI context window.
- Context Optimization Tools - Condenses memory files and tool descriptions to save space within the LLM context window.
- Context Window Management - Reduces the size of memory files and tool descriptions to maximize the available AI context window.
- Context Window Optimizations - Provides condensed formatting for server descriptions to reduce token consumption within an AI agent's context window.
- Fragment-Based Message Formats - Replaces full grammatical sentences with concise technical keywords to minimize token consumption.
- Memory Compression - Condenses project documentation and memory files into a shorter format to reduce input tokens.
- Linguistic Compression Skills - Implements a set of instructions that force AI to communicate using a simplified, fragment-based language.
- Prompt Compression - Uses system prompts to force the AI into a condensed, fragmented communication style.
- Claude Code Configurations - Optimizes the Claude Code CLI experience by compressing tool descriptions and communication styles.
- Terse Review Assistants - Provides an AI configuration for generating terse one-line pull request comments and short commit messages.
- Terse Reviews - Generates concise, one-line pull request comments to accelerate the technical review process.
- LLM Cost Management - Calculates session token usage and translates savings into real-time monetary cost reductions.
- Concise Reviews - Generates brief and direct pull request comments and commit messages to speed up technical reviews.
- AI Cost Monitoring - Tracks real-time token usage and calculates financial savings achieved through condensed communication.
- Token Usage Analytics - Calculates real-time session token consumption and lifetime cost savings based on session logs.
- Token Cost Calculators - Translates raw token counts from session logs into monetary savings using specific pricing models.
- Token Usage Logs - Monitors total tokens processed across conversations using session logs to quantify compression efficiency.
Star history
How this analysis was created: This summary and feature list were written by an AI model that read the project's README and public documentation pages. Each feature links to the documentation it came from; stars, license and language come straight from the GitHub API. The model does not read the source code, and the analysis is refreshed when the project is re-analysed. Learn more on our About page.
