why use many token when few do trick

Your AI coding agent bills by the word and writes like it knows that.
Caveman make it stop. Brain still big. Mouth small. Bill small.

See it · Install · Numbers · Skill · Proxy · Wrap · Docs · Privacy · License

See it

🗣️ Normal agent · 69 tokens	Caveman agent · 19 tokens
The reason your React component is re-rendering is likely because you're creating a new object reference on each render cycle. When you pass an inline object as a prop, React's shallow comparison sees it as a different object every time, which triggers a re-render. I'd recommend using useMemo to memoize the object.	New object ref each render. Inline object prop = new ref = re-render. Wrap in useMemo.

Same diagnosis. Same fix. Same useMemo. The only thing that died was the throat-clearing.

Code, commands, file paths, and exact error messages never get cavemanned. Only the prose around them does.

Caveman no make brain smaller. Caveman make mouth smaller.

Install

Caveman come in two sizes.

Small rock: the skill. A rule file that makes your agent answer in caveman. MIT, free forever, works in 30+ agents (Claude Code, Codex, Gemini, Cursor, Windsurf, Cline, Copilot, more). One command:

npx skills add JuliusBrussee/caveman -g

Type /caveman if your agent doesn't wake up on its own. That the whole install. One rock.

Big rock: the proxy. Runs on your machine, between your agent and the AI provider, and shrinks what the agent reads before every call. Everything it squeezes gets a backup on your disk, so the agent can always pull the original back. MIT CLI, BSL-1.1 runtime:

npm install -g @caveman-ai/cli && caveman setup --install
caveman claude        # or codex · gemini · aider · kilo · qwen · opencode · hermes · openclaw · pi

They stack. Most people start with the small rock and graduate.

curl -fsSL https://raw.githubusercontent.com/JuliusBrussee/caveman/v2.6.0/install.sh | bash

irm https://raw.githubusercontent.com/JuliusBrussee/caveman/v2.6.0/install.ps1 | iex

# Claude Code
claude plugin marketplace add JuliusBrussee/caveman && claude plugin install caveman@caveman
Gemini CLI
gemini extensions install https://github.com/JuliusBrussee/caveman
Qwen Code CLI, then its Caveman wrapper
npm i -g @qwen-code/qwen-code
caveman qwen
Codex, Cursor, Windsurf, Cline, and other skills-compatible agents
npx skills add JuliusBrussee/caveman --skill '*' -a codex --yes -g  # replace codex with your agent profile

The full 30+ agent matrix, dry runs, flags, and verification live in INSTALL.md.

The numbers

A token is what AI billing counts, roughly three quarters of a word. Your agent pays for every token it writes and every token it reads. Reading is usually the bigger bill. The skill cuts the writing. The proxy cuts the reading.

Skill: writing less

The skill asks for shorter prose. A general output-reduction percentage is not currently supported by reviewed raw results from the API benchmark in this repository. The older ten-prompt table and its fixed 65% claim have been retired.

No reviewed API benchmark result is published here yet. Run uv run python benchmarks/run.py to generate a new result, then review its raw response pairs and quality before publishing the generated table.

A separate committed evaluation snapshot compares skill output with a plain Answer concisely. control. It measures approximate output length, not correctness, billing, or performance of the current skill. Its model, date, tokenizer, and limits are recorded alongside it.

The rules add input tokens; their cost depends on which rules the host loads, when it injects them, and caching. Shorter output alone does not establish net savings. /caveman-stats reports recorded usage with savings unknown. Full accounting and workload limitations: docs/HONEST-NUMBERS.md.

Proxy: reading less

Your agent rereads logs, test output, diffs, and half your repo all day. The proxy shrinks that stream before it reaches the provider. Pinned 54-run Claude Code benchmark, provider-reported input tokens, three runs per case:

| Case | Direct Claude Code | Through caveman | Change | | ---------------------- | -----------------: | --------------: | ---------: | | CSV outlier hunt | 165,823 | 74,484 | -55.1% | | Log needle in haystack | 148,807 | 74,068 | -50.2% | | YAML config drift | 132,124 | 71,027 | -46.2% | | Test output failure | 150,377 | 108,514 | -27.8% | | Deployment JSON drift | 147,975 | 108,939 | -26.4% | | Dashboard HTML alert | 140,687 | 154,641 | +9.9% | | Total | 885,793 | 591,673 | -33.2% |

All 18 of 18 fixture answer checks passed. This does not establish quality on other tasks. These are pinned reported results; the raw harness and run artifacts are not in this checkout. Method, confidence intervals, and limits: docs/WRAP-BENCHMARK.md.

Maintainer note. The HTML row is red and it stays red. That case had no compression transform, so caveman paid its own overhead and won nothing back. The day I hide a red row is the day you should stop trusting the green ones.

Browsing too: a focused question against a 200-row table costs 121 tokens through caveman's view of the page, against 15,704 for the Playwright ARIA baseline. That's 129.8× smaller (browse/BENCHMARK.md).

The skill, unpacked

One rule file, one talking style, plus a small toolbox. /caveman lite|full|ultra|wenyan-lite|wenyan-full|wenyan-ultra sets intensity. /caveman off or normal mode turns it off.

The proxy, unpacked

One local process. Your agent talks to it, it talks to your provider. No Caveman server in the path, and your Claude Pro/Max login passes through to Anthropic untouched. Originals of everything it compresses sit in a SQLite file on your machine with a recovery handle, so the agent can always ask for the full version back.

Whole team? One container. Same proxy in your VPC, one shared token, keys stay on server. Deploy it →

Where your tokens go

Months of your agent history already sit on your disk. caveman learn reads it, locally, read-only, no account, and ranks your token sinks worst-first with a one-line fix behind each.

caveman learn             # Claude Code + Codex + Gemini CLI + opencode; aider via CAVEMAN_AIDER_ROOT
caveman learn implement   # hand the fixes to Claude Code or Codex, one diff at a time, applied only on your yes

implement re-measures after every change and reverts anything that didn't lower tokens per turn. Caveman never makes your agent dumber to make it cheaper.

More verbs

caveman explore install         # read-only FastContext subagent: finds code as path:line
caveman shrink -- pnpm test     # compress noisy command output, byte-exact recoverable
caveman browse <url>            # local Chrome over a compressed a11y tree
caveman mem remember|recall     # durable memory; mem recover <handle> = original bytes
caveman trial -- claude         # A/B a real session, then trial report
caveman toon encode|decode      # the TOON re-encoder, standalone
caveman stats                   # token history, API estimates, subscription equivalents

Pixel mode

Caveman eating its own tail. Every skill you install is prompt text your agent reloads on every call. caveman convert renders the skill body to PNG pages in place, and the model reads it as an image. On the caveman skill itself: 1,069 to 415 estimated tokens, a 61% cut.

caveman convert --dry-run        # every installed skill, with the token math, no writes
caveman convert --agent claude   # convert the profitable ones
caveman convert --revert         # byte-identical restore from SKILL.orig.md

Convert only fires when pages beat the text. Any failure leaves the skill byte-identical and names the gate that said no.

Wrap any agent

caveman turns the proxy on for good and launches the agent. caveman wrap runs one session and leaves nothing behind. It never edits your config files.

| Agent | Vendor | How it's wrapped | | -------------------- | ---------------- | ------------------------------------------------------------ | | Claude Code | Anthropic | env vars | | OpenAI Codex CLI | OpenAI | env vars (API key) · ephemeral CODEX_HOME (ChatGPT login) | | Gemini CLI | Google | env vars | | Aider | OpenAI/Anthropic | env vars | | Kilo Code | Kilo Code | KILO_CONFIG_CONTENT, your kilo.json untouched | | Qwen Code | QwenLM | ephemeral system-settings overlay, source settings untouched | | opencode | sst | inline config via env, your opencode.json untouched | | Hermes Agent | Nous Research | --provider custom + env | | OpenClaw | OpenClaw | ephemeral merged config, your config read-only | | Pi | pi.dev | bundled native extension, your ~/.pi config untouched |

The whole cave

One idea everywhere: agent do more with less.

| Repo | What it shrinks | Status | | --------------------------------------------------------------------- | -------------------------------------------------------- | ----------------- | | caveman (you here) | What the agent says (skill) and reads (proxy) | live | | caveman-browse | What the agent sees in the browser | live | | caveman-agent-sdk | What your production agent loads, calls, and spends | own repo · in dev | | cavegemma | The compression baked into weights (Gemma fine-tune) | labs | | caveman-code | The whole agent, end to end | frozen | | cavemem | What the agent remembers, across sessions | frozen | | cavekit | The build loop, spec-driven | frozen |

Frozen ones still install and work. Their best ideas moved in here.

Caveman make token small. Caveman Cloud make it provable. Local numbers are inferred, pinned benchmarks benchmark_counterfactual, neither is an invoice. Live traffic behind eval gates with signed receipts earns verified. That's Cloud. Waitlist at caveman.so

Privacy, and a small favor

Your agent still talks to the provider you chose. The skill and hooks run entirely on your machine, and nothing here needs an account.

The caveman CLI does send anonymous usage stats by default, and here's the honest why: caveman is free, one person maintains it, and those stats are how I find out which commands people actually use and which optimizations run in real workflows. That's what keeps this thing free and pointed in the right direction. Fair trade, we think.

What it sends: which commands ran, plus token counts through and cut. What it never sends: your prompts, your code, your file paths, or anything that could identify you. It tells you all this the first time you run it.

Not into it? One command and it's off forever, no hard feelings:

caveman telemetry off      # or set DO_NOT_TRACK=1

Exact network, telemetry, and storage boundaries: SECURITY.md.

License

Split license. Skill and adoption surfaces are MIT. Engine-linked runtime is BSL-1.1 source-available, not OSI Open Source before Change Date.

MIT: the skill, Agent SDK and initializer, the CLI, both client SDKs, contracts, provider catalog, extension shell, and the thin cavemem clients. Free like mammoth on open plain.

BSL-1.1: Engine, Proxy, Cache Engine, rewriter, Browse, MCP server, shrink, cavemem Go core, and shared Go platform. New Engine-linked runtime modules default to BSL-1.1. Read it, fork it, self-host it for your own first-party traffic free, production included. Each version converts to Apache-2.0 on the earlier of 2030-06-21 or four years after it ships. Hosting it for third parties needs a commercial license.

engine/pixel embeds pxpipe (MIT) plus glyph atlases derived from Spleen 5×8 (BSD-2-Clause) and GNU Unifont (dual OFL-1.1 / GPLv2-with-font-exception); its NOTICE travels with that source.

"Caveman" and the rock logo are trademarks of Julius Brussee. "Powered by Caveman" is fine when true.

Star this repo

Caveman save you token, save you money. Star cost zero. Fair trade. ⭐

_Docs: Technical manual · Install matrix · Honest numbers · Wrap benchmark · License · Contributing · Maintainer guide · Issues
MIT skill · BSL-1.1 engine. Few token. No lie.