Caveman

Cuts your token usage by about 75% by having Claude respond like a technically accurate caveman. No articles, no filler words, just the substance. It has intensity levels from lite (just removes pleasantries) to ultra (aggressive abbreviation) to wenyan modes that use classical Chinese grammar for maximum compression. Good for when you're burning through tokens on repetitive tasks or long coding sessions. It smartly breaks character for security warnings and destructive operations, then snaps back. The persistence mechanism is solid, it won't drift back to verbose mode after a few turns like some prompt hacks do. Honestly kind of hilarious to use while still being genuinely useful for API costs.

GitHub stars: 504
last updated: Jun 11, 2026

View on GitHub

Install

$ npx -y skills add hoangnguyen0403/agent-skills-standard --skill caveman --agent claude-code

Installs 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"). Technical terms exact. Code blocks unchanged. Errors quoted exact.

Pattern: [thing] [action] [reason]. [next step].

Not: "Sure! I'd be happy to help you with that. The issue you're experiencing is likely caused by..." Yes: "Bug in auth middleware. Token expiry check use < not <=. Fix:"