caveman

🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman

Frozen — cross-agent persistent memory for coding assistants. Still works; the compressed-memory core now ships inside JuliusBrussee/caveman.

genshijin 原始人 🗿| Claude Code / Codex等AIエージェント 向け超圧縮コミュニケーションスキル。caveman の日本語版をベースに、日本語特有の冗長表現に最適化。

Your friendly, blazing-fast, self-hostable AI gateway

LoRA fine-tune Gemma 4 31B to speak caveman-mode natively. Style: github.com/JuliusBrussee/caveman

🪨 why install when copy do trick - Simple Caveman prompt files for GitHub Copilot without install.

OpenCode package for Caveman: terse AI responses, slash commands, compact reviews, commit messages, and markdown memory compression.

We have caveman system prompts and skills for AI models to reduce token use, why not try bake it into the model itself with fine-tuning?

🔥 Save up to 96% tokens — more than Caveman (65%) or RTK (80%). AI coding agent: chat, map, edit, multi-agent. Single Rust binary.

Run many Codex & Claude agents in parallel without them overwriting each other. Isolated worktrees, file locks, PR-only merges. Auto-wires Oh My Codex, Oh My Claude, OpenSpec, and Caveman in every worktree.

Caveman output style for Claude Code: 40% fewer output tokens, always-on formatting

Multi-agent orchestration plugin for OpenCode. 10 configurable model slots, 30 hooks, memory search, LSP diagnostics, agent-browser integration, completion controller, auto dream/distill memory consolidation. One prompt to set up and run.

Multi-agent skill for faster workflows

Caveman prompting, measured. A two-channel evaluation protocol scoring what input and output compression actually cost LLMs in dollars, accuracy, and surface-text fidelity across seven models and five benchmarks.

websocket-driver caveman chat example.

✂️ Token-saving OMP add-ons: terse caveman replies, compact RTK shell output, lean Ponytail code calls — one-command combo presets.

Plugin для Claude Code що вмикає печерний режим — видаляє воду, зберігає суть. −65–90% токенів без втрати точності. Побудований для українських розробників, розуміє обидві мови.

Comprehensive Cursor skill for Agent token efficiency: Ponytail, Caveman, Headroom and workflow patterns

One command turns any repo into a governed, AI-driven delivery pipeline for Claude Code: Beads task memory, code graph + semantic RAG, agents, Ollama, team sync, token savings.