caveman CLAUDE.md
From JuliusBrussee
README = product front door. Non-technical people read it to decide if caveman worth install. Treat like UI copy.
Instructs Claude to treat caveman's README as user-facing product copy, not internal docs.
Use it when
- Editing README.md, INSTALL.md, or the What You Get feature table
- Adding a new agent to the install table (needs matching detail block)
- Changing skill behavior in skills/*/SKILL.md and updating docs to match
- Verifying benchmark numbers before quoting them in README
Skip it if
- Not maintaining the caveman project's own README/docs
- You want general README-writing advice, not caveman's specific voice/layout rules
- You don't need the caveman-specific file-ownership map (SKILL.md, rules/, agents/)
Facts
- Repository
- JuliusBrussee/caveman
- Status
- Actively maintained
- Last commit
- Source file
- CLAUDE.md
Source preview
The instructions Claude Code reads when this claude.md runs.
# CLAUDE.md — caveman
## README is a product artifact
README = product front door. Non-technical people read it to decide if caveman worth install. Treat like UI copy.
**Rules for any README change:**
- Readable by non-AI-agent users. If you write "SessionStart hook injects system context," invisible to most — translate it.
- Keep Before/After examples first. That the pitch.
- Install table always complete + accurate. One broken install command costs real user.
- What You Get table must sync with actual code. Feature ships or removed → update table.
- Preserve voice. Caveman speak in README on purpose. "Brain still big." "Cost go down forever." "One rock. That it." — intentional brand. Don't normalize.
- Benchmark numbers from real runs in `benchmarks/` and `evals/`. Never invent or round. Re-run if doubt.
- Adding new agent to install table → add detail block in `<details>` section below.
- Readability check before any README commit: would non-programmer understand + install within 60 seconds?
---
## Project overview
Caveman makes AI coding agents respond in compressed caveman-style prose — cuts 65% output tokens (measured), full technical accuracy. Ships as Claude Code plugin, Codex plugin, Gemini CLI extension, agent rule files for Cursor, Windsurf, Cline, Copilot, 40+ others via `npx skills`.
---
## What lives where
Post-cleanup layout. Sources of truth at the topView full source on GitHub →Other claude.md files
ECC CLAUDE.md
★ 229,918This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
affaan-mupdated 15d agoMITui-ux-pro-max-skill CLAUDE.md
★ 105,847This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
nextlevelbuilderupdated 14d agoMITservers CLAUDE.md
★ 88,502This file provides guidance to Claude Code when working with code in this repository.
modelcontextprotocolupdated 19d agoNOASSERTIONclaude-mem CLAUDE.md
★ 87,329Claude-mem is a Claude Code plugin providing persistent memory across sessions. It captures tool usage, compresses observations using the Claude Agent SDK, and injects relevant context into future ses
thedotmackupdated 13d agoApache-2.0agent-skills CLAUDE.md
★ 78,448This is the agent-skills project — a collection of production-grade engineering skills for AI coding agents.
addyosmaniupdated 16d agoMITUnderstand-Anything CLAUDE.md
★ 74,200An open-source tool combining LLM intelligence + static analysis to produce interactive dashboards for understanding codebases.
Egonex-AIupdated 14d agoMIT