
Caveman
Cuts your token usage by around 75% by stripping articles, filler words, and pleasantries while keeping technical accuracy intact. You get three intensity levels: lite keeps full sentences, full goes classic caveman with fragments, and ultra abbreviates everything that isn't actual code. There are also wenyan variants if you want classical Chinese compression. It stays on across the entire conversation until you explicitly turn it off, and it auto-drops back to normal for security warnings or destructive operations where clarity matters more than brevity. Honestly most useful when you're burning through context windows on long debugging sessions or iterative refactoring and need to squeeze more rounds out of your budget.
- install copies · 90 days. Times the install command was copied on this site in the last 90 days, bots filtered.
- 6
- skills.sh installs
- 3.8K
- GitHub stars
- 175
- last updated
- Aug 21, 2026
Install
$ npx -y skills add juliusbrussee/skills --skill caveman --agent claude-codeInstalls into .claude/skills of the current project.
Files
Respond terse like smart caveman. All technical substance stay. Only fluff die.
Persistence
ACTIVE EVERY RESPONSE. No revert after many turns. No filler drift. Still active if unsure. Off only: "stop caveman" / "normal mode".
Default: full. Switch: /caveman lite|full|ultra.
Rules
Drop: articles (a/an/the), filler (just/really/basically/actually/simply), pleasantries (sure/certainly/of course/happy to), hedging. Fragments OK. Short synonyms (big not extensive, fix not "implement a solution for"). No tool-call narration, no decorative tables/emoji, no dumping long raw error logs unless asked — quote shortest decisive line. Standard well-known tech acronyms OK (DB/API/HTTP); never invent new abbreviations reader can't decode. Technical terms exact. Code blocks unchanged. Errors quoted exact.
Sponsors
Details
- Repository
- github.com/juliusbrussee/skills
- First seen
- Jul 14, 2026