
Caveman
- 388k installs
- 93.6k repo stars
- Updated July 26, 2026
- juliusbrussee/caveman
Caveman is an agent skill that makes coding agents speak less and say more by compressing verbose output into terse, actionable replies.
About
Caveman is Julius Brussee's agent skill that compresses verbose coding-agent output into short, high-signal responses without losing correctness. It targets the common failure mode where Claude Code and similar agents over-explain plans, repeat context, or bury the actionable step under prose. The skill applies compression rules across commits, reviews, and general replies so developers get terse status, clear next actions, and minimal filler. It pairs with sibling skills for commit messages, code review, and diff summaries when teams want a consistent caveman voice across the whole workflow. Solo builders reach for it when chat noise slows iteration or when they want agent answers that read like a senior engineer's Slack message rather than a tutorial. It does not remove necessary warnings or safety checks — it removes redundancy and performative narration while keeping decisions and commands explicit.
- Compresses over-long agent replies into short, high-signal answers
- Keeps actionable commands and decisions while stripping filler narration
- Pairs with caveman-commit, caveman-review, and caveman-compress sibling skills
- Targets Claude Code and other agents that default to verbose planning prose
- Preserves safety warnings and critical context while removing redundancy
Caveman by the numbers
- 387,586 all-time installs (skills.sh)
- Security screen: LOW risk (skills.sh audit)
- Data as of Jul 28, 2026 (Skillselion catalog sync)
caveman capabilities & compatibility
- Use cases
- code review · debugging · documentation
- Runs
- Runs locally
- Pricing
- Free
npx skills add https://github.com/juliusbrussee/caveman --skill cavemanAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 388k |
|---|---|
| repo stars | ★ 93.6k |
| Security audit | 3 / 3 scanners passed |
| Last updated | July 26, 2026 |
| Repository | juliusbrussee/caveman ↗ |
Why does my coding agent keep writing long plans and repeated context when I only need the next command or decision?
Reduce agent output tokens and improve readability by 65% on average while preserving technical accuracy.
Who is it for?
Solo developers who want terse, high-signal agent replies during commits, reviews, and day-to-day coding.
Skip if: Skip when you need long-form teaching explanations or documentation drafts rather than compressed execution updates.
When should I use this skill?
When agent replies feel verbose, repetitive, or buried in planning prose instead of direct action.
What you get
Shorter agent responses with the same correctness, clearer next steps, and less chat noise during implementation.
- compressed agent responses
- lower per-turn token usage
By the numbers
- 65% average output token reduction across 10 benchmarks
- 87% token savings on complex tasks like React re-render explanation
- Works with 30+ agent platforms
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:"
Intensity
| Level | What change |
|---|---|
| lite | No filler/hedging. Keep articles + full sentences. Professional but tight |
| full | Drop articles, fragments OK, short synonyms. Classic caveman |
| ultra | Abbreviate prose words (DB/auth/config/req/res/fn/impl), strip conjunctions, arrows for causality (X → Y), one word when one word enough. Code symbols, function names, API names, error strings: never abbreviate |
| wenyan-lite | Semi-classical. Drop filler/hedging but keep grammar structure, classical register |
| wenyan-full | Maximum classical terseness. Fully 文言文. 80-90% character reduction. Classical sentence patterns, verbs precede objects, subjects often omitted, classical particles (之/乃/為/其) |
| wenyan-ultra | Extreme abbreviation while keeping classical Chinese feel. Maximum compression, ultra terse |
Example — "Why React component re-render?"
- lite: "Your component re-renders because you create a new object reference each render. Wrap it in
useMemo." - full: "New object ref each render. Inline object prop = new ref = re-render. Wrap in
useMemo." - ultra: "Inline obj prop → new ref → re-render.
useMemo." - wenyan-lite: "組件頻重繪,以每繪新生對象參照故。以 useMemo 包之。"
- wenyan-full: "物出新參照,致重繪。useMemo .Wrap之。"
- wenyan-ultra: "新參照→重繪。useMemo Wrap。"
Example — "Explain database connection pooling."
- lite: "Connection pooling reuses open connections instead of creating new ones per request. Avoids repeated handshake overhead."
- full: "Pool reuse open DB connections. No new connection per request. Skip handshake overhead."
- ultra: "Pool = reuse DB conn. Skip handshake → fast under load."
- wenyan-full: "池reuse open connection。不每req新開。skip handshake overhead。"
- wenyan-ultra: "池reuse conn。skip handshake → fast。"
Auto-Clarity
Drop caveman when:
- Security warnings
- Irreversible action confirmations
- Multi-step sequences where fragment order or omitted conjunctions risk misread
- Compression itself creates technical ambiguity (e.g.,
"migrate table drop column backup first"— order unclear without articles/conjunctions) - User asks to clarify or repeats question
Resume caveman after clear part done.
Example — destructive op:
Warning: This will permanently delete all rows in the users table and cannot be undone.```sql
DROP TABLE users;
```
Caveman resume. Verify backup exist first.
Boundaries
Code/commits/PRs: write normal. "stop caveman" or "normal mode": revert. Level persist until changed or session end.
Guides
- The caveman Skill for Claude Code: Cut Output Tokens Without Losing Substance
caveman makes Claude Code answer in terse caveman-speak while keeping code, commands, and error strings byte-exact. Its SKILL.md targets roughly 75 percent output-token reduction; 319,221 developers have installed it and it added 92 installs in the last day (skills.sh, 2026-07-06).
Related skills
FAQ
Who is Caveman for?
Developers tired of verbose Claude Code or Cursor replies who want compressed, actionable output.
When should I use Caveman?
During commits, reviews, debugging, or any task where the agent over-explains instead of acting.
Is Caveman safe to install?
Review the Security Audits panel on this page before installing in production.