🪨 why use many token when few token do trick. Viral skill + proxy for coding agents that cuts 65% of tokens by talking like a caveman.
★ 120,419⑂ 7,029👁 259⑃ 813● 638▤ 56
+185
日增长
+1,940
周增长
+9,211
月增长
120,419
stars
7,029
forks
259
watchers
813
prs
638
issues
56
releases
README
New: Caveman Code. Free. The most efficient coding agent. Same model, same work, smaller bill. 21% cheaperthan CodexTerminal-Bench 4, per solved task 14% cheaperthan Claude CodeTerminal-Bench 4, 97 vs 83 solved 76% cheaperthan Claude Codefull workday, 4 connected apps Same model on both sides: GPT-6.1 Sol against Codex, Opus 5.5 against Claude Code. Our own runs, November 2026, at list prices. How we measured Every coding agent in one window. Free on the subscriptions and keys you already have. Download Caveman Code for Mac · Read the launch post --- why many token when few do trick Caveman make your AI agent say less and read less. Code stay exact. Brain still big. 33.2% fewerinput tokens, whole sessions54 Claude Code runs, 18 of 18 answers right 54% cheaperwith Auto, same score85 coding tasks vs Opus 5.5 at xhigh, interim 397 → 76tokens for one test runBlocks vs | tail -20 1.4 to 2.4× cheaperwith caveman-style outputAdobe Research, eight models Cited by Adobe Research · A/B tested by JetBrains · Remade for Elasticsearch on Elasticsearch Labs 1 on Hacker News · 1 on GitHub Trending · "No way this actually works." ThePrimeagen Install · Six rocks · Proof · Say less · Read less · Auto · Blocks · Commands · Caveman Cloud --- Normal agent · 63 tokens Caveman agent · 20 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, so React re-renders. Wrap the prop in useMemo. Same fix. 63 token become 20. Brain still big. Pick your club: | Skill | Same answer | Tokens | |---|---|---:| | /caveman | New object ref each render, so React re-renders. Wrap the prop in useMemo. | 20 | | /ultracave | Inline object prop, new ref, re-render. useMemo. | 14 | | /megacave | 新參照致重繪。useMemo。 | 13 | Token counts: tiktoken o200k. Install bash npx @caveman-ai/cli@latest One command. Mac, Linux, Windows. Caveman find your agents and your logins, show what it change. Press Enter. Done. Needs Node.js 22.13+. After that, plain claude stays caveman'd. So do Codex, Gemini CLI, OpenCode, and 30+ more. It runs on your machine, with your keys and your Claude Pro/Max login. One rock. That it. Had caveman before 4.0? Say @latest, or the old copy on your machine wins. Already set up? caveman setup opens the screen again. Only want shorter answers? Get just the skill. Six rocks, one install New in 4.0: caveman is six parts. All on after install. Each has its own switch. | Rock | What it does | Measured | |---|---|---| | Say less | Agent answers short. Code stays exact | 63 → 20 tokens, same fix | | Read less | Logs, JSON, CSV, test output shrink before the model reads them | 33.2% fewer input tokens per session | | Auto · new | Picks
Top issues
#1318Ultracave ignored on Opus 5.5 despite #929, hooks fire every turn12 comments · 2026-10-17#1482middleware-canary: TypeScript adapters fail against latest framework releases11 comments · 2026-11-06#1494middleware-python nightly canary is failing9 comments · 2026-11-07#1020print mode (`claude -p`): eligible subscription requests never reach the engine, while interactive sessions compress normally7 comments · 2026-09-09#1225agent-drift: pi — installed 1.5.2 exceeds pinned 1.0.46 comments · 2026-10-08#863support for nix package manager6 comments · 2026-08-18#1285Don't work in the latest version of Antigravity CLI.5 comments · 2026-10-15#1258Visual Studio Copilot: caveman ultra recognized - not used4 comments · 2026-10-13
Discussions
#1521Turkish video about caveman (with English subtitles)0 comments · 2026-11-12#1520🏆 caveman earned Repocharts Silver: Top gainer in Go0 comments · 2026-11-12#1469Unclear on how to use with VS Code Copilot1 comments · 2026-11-05#1450Don't see Haiku model in claude code1 comments · 2026-11-02#1428Idea: Combining Caveman's prompt compression with a local FinOps proxy (hard budget caps & input savings)0 comments · 2026-10-29#1347caveman is #1 in Skills, plugins and prompts on Archestack (badge, if you want it)0 comments · 2026-10-21#1197Why caveman installer does not deploy $HOME/.copilot/copilot-instructions.md?1 comments · 2026-10-06#1179Overalls: Caveman's mode as a dropdown in a status band0 comments · 2026-10-03
Latest release
Caveman 4.1.0 v4.1.0 2026-11-08Caveman 4.1.0 is a maintenance release: bug fixes across compress, OpenCode, Hermes, Pi, the proxy and the hooks, a recovery store that no longer fills up and stops compressing, and a few small switches people asked for. caveman-compress now refuses a file with a key or token in it, and installs from skill hubs again. Action needed - forward_headers under a providers. block now stops the proxy from starting. Before, it was silently ignored: no headers were forwarded and the only symptom was an upstream 401. Configure that upstream as a compat mount instead: compat. with base_url and forward
相关报道 (8)
cyberqGitHub 趨勢周報 Vol.14,DeepSeek V4 Flash 適合本地端,Caveman 生態補上 AI Agent 的記憶與規格化流程2026-05-11cyberqCaveman 幫 AI 開發者暴力省下 75% Token 成本2026-05-11Pasquale PillitteriCaveman、Ponytail、rtk、pxpipe 与 headroom:削减 Claude Code 令牌2026-07-11Pasquale PillitteriClaude Code 令牌: 节省高达 90% 的 10 个 GitHub 仓库2026-04-21Pasquale PillitteriClaude Code Caveman Mode:削减 Token 而不破坏代码的 Skill2026-04-15智源社区省token神器3天狂揽4.1k星!19岁小哥开发,信息无损最高省87%2026-04-08虎嗅Anthropic发布最强AI模型Claude Mythos,API价格高达125美元/百万Token2026-04-08新浪财经逼AI当山顶洞人!Claude防话痨插件爆火,网友:受够了AI废话2026-04-06
News & coverage (13)
Hacker NewsCaveman: Why use many token when few token do trick▲ 904 · 366 comments · 2026-04-05Hacker NewsRespond terse like smart caveman. All technical substance stay. Only fluff die▲ 2 · 0 comments · 2026-09-10Hacker NewsLocal SLM as a compression layer for cloud API calls▲ 2 · 0 comments · 2026-04-25Hacker NewsCaveman – why use many token when few do trick▲ 2 · 0 comments · 2026-04-16techtimesJetBrains Tests Caveman Token Skill on 86 Real Tasks: Savings Hit 9%, Not 65%2026-07-16aiweekly'Caveman' plugin trims Claude and Codex output to cut AI bills2026-06-30opentools.aiCaveman Claude Code Skill2026-06-05moneywise.com'Caveman' plugin promises to cut AI costs by 75%2026-07-07Pasquale PillitteriHeadroom vs Caveman: Compress Input or Cut Output2026-06-20Pasquale PillitteriClaude Code Caveman Mode: the Skill That Cuts Tokens Without Breaking Code2026-04-14