Discover the toolingbehind modern intelligence
The operating system for AI infrastructure discovery. Search a live map of MCP servers, AI agents, LLM tools, automation systems, and developer infrastructure.
caveman
🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman
66.3kstars
3.7kforks
JavaScript
aianthropiccavemanclaudeclaude-codellmmemeprompt-engineeringskilltokens
Quality signals
ActiveNo testsNo CIThin docsNo licensePopular
AI summary
Auto-generated overview of what this skill does and when to reach for it.
A Claude Code skill that rewrites AI agent replies in caveman-speak, cutting output tokens by 65% without changing technical content.
Use this when
- You’re tired of verbose AI replies eating your token budget
- You want faster, punchier answers from coding agents
- You’re debugging and need concise error messages
- You’re reviewing PRs and want tight one-line comments
- You’re using Claude Code, Cursor, or 30+ other agents
Don't use when
- When you need detailed explanations for learning or documentation
- When the agent’s verbosity is actually helpful for context
- When working with non-technical users who need plain English
Capabilities
- Rewrites agent replies in caveman-style (lite, full, ultra, wenyan modes)
- Compresses commit messages to ≤50 chars with /caveman-commit
- Generates one-line PR review comments with /caveman-review
- Shows token savings stats with /caveman-stats
- Compresses memory files (e.g., CLAUDE.md) to cut input tokens
- Acts as MCP middleware to shrink tool descriptions
InstallSKILLInstall as a Claude Code skill
Universal installANY TOOL
$ npx skillcity@latest JuliusBrussee/cavemanAuto-detects the skill or MCP server in this repo and installs it for Claude Code, Codex, Cursor, Cline, OpenCode, Gemini CLI, Anti-Gravity, OpenClaw and Hermes.
Requires Node.js 18+ (ships with npx).
Or install manually
# Install into ~/.claude/skills/ mkdir -p ~/.claude/skills/caveman curl -fsSL https://raw.githubusercontent.com/JuliusBrussee/caveman/main/SKILL.md -o ~/.claude/skills/caveman/SKILL.md
Loading README…
Ask a simulated agent to run this skill. It loads the real SKILL.md and reasons over your task with NVIDIA.
caveman — sandboxAI sandbox
Pick a quick prompt below or type your own — watch the agent execute this skill step by step.
guest@skillcity:~$Try a prompt