Caveman: cut your agent's token use by talking short

Try it in 60 seconds

npx @caveman-ai/cli@latest
npx skills add JuliusBrussee/caveman -g
claude plugin marketplace add JuliusBrussee/caveman && claude plugin install caveman@caveman

Install command copied from the project README. Check the README for the latest version.

What it does

A skill plus proxy that makes coding agents reply in terse 'caveman' style, cutting tokens while keeping code exact.

Caveman make your AI agent say less and read less. Code stay exact. Brain still big.

33.2% fewer input tokens, whole sessions 54 Claude Code runs , 18 of 18 answers right 54% cheaper with Auto, same score 85 coding tasks vs Opus 5.5 at xhigh, interim 397 → 76 tokens for one test run Blocks vs | tail -20 1.4 to 2.4× cheaper with caveman-style output Adobe Research , eight models

Quoted from the README of JuliusBrussee/caveman.

Inside the repo

Step by step

1. Open the GitHub repo and read the README.
2. Run the install command below in your terminal (or copy the skill folder into ~/.claude/skills/ for Claude Code, ~/.codex/skills/ for Codex).
3. Restart your agent so it picks up the skill.
4. Ask your agent to use it on a small test task first.

Watch out for

Docs, demos & materials

Sources & credit

All credit to the original authors. We rewrite in our own words and link back. Something wrong? Report it.