juliusbrussee caveman
COMMUNITYLABSCO SUMMARY
The package is caveman, a Claude Code / Codex / Gemini / Cursor skill (plus 30+ other agents via its own installer) that swaps an agent's normal explanatory prose for compressed "caveman-speak" while keeping code, commands and error text byte-for-byte exact. All 7 skills here belong to that one idea: caveman itself is the toggle, with six levels from lite to a classical-Chinese wenyan mode, sticky for the session; caveman-commit, caveman-compress and caveman-review apply the same compression to commit messages, memory files like CLAUDE.md, and PR review comments; caveman-stats reports the session's measured savings; and cavecrew is a set of three subagents (investigator, builder, reviewer) that delegate the same way to keep the main context window lighter.
This suits someone reading a lot of agent output during a session — coding work in Claude Code, Codex or Gemini CLI, where terse, code-accurate answers speed up iteration. It is not aimed at anyone who needs the agent's longer explanations for onboarding or an unfamiliar codebase, and levels above lite intentionally drop connective reasoning from the visible answer, not just filler words.
READ THE FULL ANALYSIS
The 65% figure is the author's own benchmark, and the README states its own limits in the same breath. It counts output tokens only; input and reasoning tokens are untouched, and the skill's instructions add roughly 1,000-1,500 input tokens per turn — so on an already-terse workload the net effect can go negative rather than positive. The 65% average itself comes from 10 prompts the author ran and committed to the repo's benchmarks/ folder, not from an independent test.
What you install here is one piece of a five-repository family, and only this piece. The README promotes caveman-code (a full terminal coding agent), cavemem, cavekit and a Gemma fine-tune called cavegemma, plus a separate caveman-shrink npm package and a second skills repository, JuliusBrussee/skills, with four unrelated skills of its own (grill-me, interface-kit, junior-to-senior, loop-factory). None of those are part of the 7 skills counted on this page.
Our source README was truncated at 12KB of a 14.9KB file, cut off partway through a "Caveman 2" section that describes a forthcoming, more rigorously measured successor as not yet shipped — today's /caveman-stats numbers remain the author's own local estimates.
ALSO IN THIS PACKAGE
caveman
Ultra-compressed communication mode. Shortens prose while preserving technical terms, code, commands, and exact errors.
cavecrew-builder
Surgical 1-2 file edit. Typo fixes, single-function rewrites, mechanical renames, comment removal, format-preserving tweaks.
cavecrew-investigator
Read-only code locator. Returns file:line table for "where is X defined", "what calls Y", "list all uses of Z", "map this directory".
cavecrew-reviewer
Diff/branch/file reviewer. One line per finding, severity-tagged, no praise, no scope creep.
WHAT'S INSIDE
7 showing · 7 totalNothing else to set up — install it and go.
caveman-commit
Writes the short note that goes with a saved change to code: one line saying what changed, and a paragraph underneath only when the reason is not obvious from the change itself.
caveman-review
Feedback on someone's code written one line per problem — where it is, what is wrong, what to do about it — instead of a polite paragraph that buries the point.
caveman
A mode that makes the assistant stop padding its answers — replies come back clipped and telegram-short for the rest of the session, in English or in classical Chinese.
caveman-help
The cheat sheet for the whole caveman family — every mode, the command that switches it on, and where to change which one you start in — printed without actually switching anything.
cavecrew
Three helpers you can hand a small job to — find where something lives in the code, make a one-file change, check a change for mistakes — and each answers in shorthand so the main conversation does not fill up.
caveman-compress
An assistant re-reads your standing instructions on every single turn and you pay for them every time — this rewrites that file in the same clipped shorthand so it costs a fraction, and keeps the readable original somewhere safe outside your project.
caveman-stats
Shows how much text this session has actually run through the model, counted from the record rather than estimated, and it refuses to turn that into a 'you saved 30%' claim because nothing measured the alternative.
HOW TO GET IT
npx skills add juliusbrussee/cavemannpx skills add juliusbrussee/caveman --skill <name> --full-depthPick the skill name from the Skills tab — each entry there installs independently.
/plugin marketplace add juliusbrussee/caveman/plugin install caveman@cavemanTyped inside the agent's own prompt, not in a terminal. The marketplace is called caveman, which is the part after the @.