By glitchwerks
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Surgical 1-2 file edit. Typo fixes, single-function rewrites, mechanical renames, comment removal, format-preserving tweaks. Hard refuses 3+ file scope. Returns caveman diff receipt. Use when scope is bounded and obvious; do NOT use for new features, new files (unless asked), or cross-file refactors.
Read-only code locator. Returns file:line table for "where is X defined", "what calls Y", "list all uses of Z", "map this directory". Output is caveman-compressed so the main thread eats ~60% fewer tokens than vanilla Explore. Refuses to suggest fixes.
Diff/branch/file reviewer. One line per finding, severity-tagged, no praise, no scope creep. Output format `path:line: <emoji> <severity>: <problem>. <fix>.` Use for "review this PR", "review my diff", "audit this file". Skips formatting nits unless they change meaning.
Decision guide for delegating to caveman-style subagents. Tells the main thread WHEN to spawn `cavecrew-investigator` (locate code), `cavecrew-builder` (1-2 file edit), or `cavecrew-reviewer` (diff review) instead of doing the work inline or using vanilla `Explore`. Subagent output is caveman-compressed so the tool-result injected back into main context is ~60% smaller — main context lasts longer across long sessions. Trigger: "delegate to subagent", "use cavecrew", "spawn investigator/builder/reviewer", "save context", "compressed agent output".
Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman-compress FILEPATH or "compress memory file"
Quick-reference card for all caveman modes, skills, and commands. One-shot display, not a persistent mode. Trigger: /caveman-help, "caveman help", "what caveman commands", "how do I use caveman".
Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A minimized, Claude Code-only extraction of the caveman toolkit by Julius Brussee. Strips all non-Claude runtimes and external dependencies down to the surface that matters: terse compressed responses in Claude Code, with no npm install, no pip install, no API keys beyond what you already have.
Security-audited for work-machine use. See docs/security-audit.md.
Activates "caveman mode" on every session start. Claude replies with terse, technically accurate output — dropping articles, filler, pleasantries, and hedging — while keeping code blocks, error strings, paths, and commands exactly as-is.
Typical output token reduction: 65–75%. Technical accuracy: unchanged. Code/commits/PRs are always written in normal prose.
Safety valve: Claude automatically drops caveman for security warnings, irreversible-action confirmations, and any step where fragment order could be misread.
| Command | What it does |
|---|---|
/caveman | Activate full mode (default) |
/caveman lite | Drop filler; keep sentence structure and articles |
/caveman full | Drop articles, filler, pleasantries, hedging; fragments OK |
/caveman ultra | Maximum compression; abbreviate prose words; arrows for causality |
/caveman wenyan-lite | Classical Chinese register, light compression |
/caveman wenyan | Full 文言文; 80–90% character reduction |
/caveman wenyan-ultra | Extreme classical compression |
/caveman off | Deactivate for the session |
normal mode | Natural language — same as /caveman off |
/caveman-help | One-shot reference card |
/caveman-compress <file> | Rewrite a Markdown file into caveman prose (input-token saver) |
Rewrites a Markdown memory file (e.g. CLAUDE.md, notes, todos) into compressed caveman prose so every future session that loads that file pays fewer input tokens.
<name>.original.md backup before overwritingclaude --print session so the current session never ingests the file contents.env, credentials.*, id_rsa, *.pem, etc.) before any readclaude CLI — the same endpoint Claude Code already uses. Do not point it at restricted content.Three terse subagent presets that return compressed output. Because subagent tool results are injected verbatim into main context, compressed output extends how far the context budget goes across a long session.
| Agent | Purpose |
|---|---|
cavecrew-investigator | Read-only code locator — find symbols, callers, usages |
cavecrew-builder | Surgical editor for 1–2 file changes when the site is already known |
cavecrew-reviewer | Diff/file reviewer; returns structured findings, not prose |
Use /cavecrew (the skill) to decide when delegation makes sense vs. working in the main thread.
Default is full. Override via environment variable or config file.
Environment variable (highest priority):
export CAVEMAN_DEFAULT_MODE=lite
Config file:
// ~/.config/caveman/config.json (macOS/Linux)
// %APPDATA%\caveman\config.json (Windows)
{ "defaultMode": "lite" }
Set "off" to disable auto-activation on session start while keeping manual /caveman available.
Resolution order: CAVEMAN_DEFAULT_MODE env var > config file > full.
None beyond what Claude Code already ships.
/caveman-compress additionally requires the claude CLI on PATH (already present if you are using Claude Code)npm install, no pip install, no additional API keysOption 1 — local marketplace (recommended):
/plugin marketplace add /i/ai/claude/mini-cavemen/mini-caveman
/plugin install mini-caveman
Option 2 — plugin dir flag at launch:
claude --plugin-dir /i/ai/claude/mini-cavemen/mini-caveman
After install, start a new session and say "talk like caveman" or run /caveman. Stop with "normal mode".
npx claudepluginhub glitchwerks/mini-caveman --plugin mini-cavemanClaude Code efficiency and hygiene toolkit. Surfaces token spend across the 5h / 7d / Sonnet-7d billing windows with per-model / per-agent / per-skill attribution, and audits your effective Claude Code configuration (custom + plugin agents and skills) for overlap and conflicts. Ships an interactive dashboard, a conversational usage-analysis skill, and a structured config-audit skill.
Helps Claude make deterministic, auditable choices about which agent and skills to use for a given task — replacing prose-scanning agent/skill selection with a typed scoring kernel.
GitHub workflow skills for Claude Code - issue creation, PR review triage, backlog summaries/quick-wins, and GitHub Actions authoring.
Claude harness - A harness for solo developers (Vibecoders) to handle full-cycle contract development.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Plugin-safe Claude Code distribution of Antigravity Awesome Skills with 1,561 supported skills.