By preludium
Personal skill library — curated Claude Code skills for compressed communication, code review, and development workflows.
Designs systems and captures decisions grounded in the existing codebase. Use for system design, architectural trade-offs, or ADRs.
Root cause analysis. Use when something is broken and the cause is unclear, or when you need a precise diagnosis before writing a fix.
Writes code and tests following existing codebase patterns. Use for new features, bug fixes, or refactors.
Coordinates specialist agents for tasks that span multiple domains. Invoke explicitly: "use the orchestrator", "orchestrate this", or when asked whether to use it.
Reviews diffs and PRs, producing copy-paste-ready comments. Use when reviewing a diff, PR, or set of changed files.
Ultra-compressed commit message generator. Cuts noise from commit messages while preserving intent and reasoning. Conventional Commits format. Subject ≤50 chars, body only when "why" isn't obvious. Use when user says "write a commit", "commit message", "generate commit", "/commit", or invokes /caveman-commit. Auto-triggers when staging changes.
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 code review comments. Cuts noise from PR feedback while preserving the actionable signal. Each comment is one line: location, problem, fix. Use when user says "review this PR", "code review", "review the diff", "/review", or invokes /caveman-review. Auto-triggers when reviewing pull requests.
Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), 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.
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"
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.
Personal Claude Code plugin — curated skills for compressed communication, code review, and structured reasoning.
Add to ~/.claude/settings.json:
{
"extraKnownMarketplaces": {
"claude-kit": {
"source": {
"source": "github",
"repo": "preludium/claude-kit"
}
}
},
"enabledPlugins": {
"preludium-skills@claude-kit": true
}
}
Then run /reload-plugins in Claude Code.
Token-efficient communication. Cuts filler, keeps technical accuracy.
| Skill | Trigger | What it does |
|---|---|---|
caveman | /caveman, "less tokens", "be brief" | Ultra-compressed responses. ~75% token reduction. Levels: lite / full (default) / ultra / wenyan-* |
caveman-commit | /caveman-commit, "write a commit" | Terse Conventional Commits. Subject ≤50 chars. Body only when why isn't obvious |
caveman-review | /caveman-review, "code review" | One-line PR comments: L42: 🔴 bug: problem. fix. |
caveman-help | /caveman-help, "caveman help" | Quick reference card for all caveman commands |
compress | /caveman:compress <file> | Compress CLAUDE.md / memory files into caveman format. Backs up original as FILE.original.md |
Caveman levels:
| Level | Style |
|---|---|
lite | No filler. Keep articles + full sentences |
full | Drop articles, fragments OK, short synonyms |
ultra | Abbreviations, arrows for causality (X → Y), one word when enough |
Switch levels: /caveman lite · /caveman ultra · etc.
Deactivate: say stop caveman or normal mode.
| Skill | Trigger | What it does |
|---|---|---|
grill-me | "grill me" | Relentless interviewer. Walks every branch of a design decision tree, one question at a time, exploring the codebase when answers can be found there |
grill-with-docs | "grill me on this", "stress-test against docs" | Like grill-me but challenges your plan against the existing domain model, sharpens terminology, and updates CONTEXT.md / ADRs inline as decisions crystallise |
the-fool | "challenge this", "devil's advocate", "red team" | Structured critical reasoning across 5 modes: Socratic questioning, Dialectic synthesis, Pre-mortem, Red team, Evidence audit. Steelmans first, then challenges |
The Fool modes:
| Mode | What it does |
|---|---|
| Expose my assumptions | Socratic probing — surfaces what's being taken for granted |
| Argue the other side | Hegelian dialectic — steelmans the opposing position |
| Find failure modes | Pre-mortem — ranks how this fails, with mitigations |
| Attack this | Red team — adversary profiles, attack vectors, defenses |
| Test the evidence | Falsificationism — audits whether evidence supports the claim |
| Skill | Trigger | What it does |
|---|---|---|
karpathy-guidelines | Auto-triggers when writing/reviewing/refactoring code | Guardrails against common LLM coding mistakes: think before coding, simplicity first, surgical changes, goal-driven execution with verifiable success criteria |
review-diff | /review-diff, "review my changes", "review this diff" | Runs git diff main...HEAD, classifies findings as security / bug / risk / nit, outputs copy-paste-ready PR comments. Jargon-free, junior-readable, suggestion blocks for nits. |
| Skill | Trigger | What it does |
|---|---|---|
mentor-mode | "teach me X", "I want to understand", "I'm learning X", "help me figure this out" | Switches Claude from code executor to programming mentor. Guides with questions, hints progressively, checks understanding. Calibrates to beginner / intermediate / advanced. |
| Skill | Trigger | What it does |
|---|---|---|
handoff | "handoff", "summarise for next session" | Compacts the current conversation into a handoff document saved to the OS temp dir. Suggests skills for the next agent. Redacts sensitive info. Accepts an argument describing what the next session will focus on. |
| Agent | Trigger | What it does |
|---|---|---|
reviewer | "review this PR", "review the diff", /reviewer | Produces copy-paste-ready PR comments. Each comment is brief, jargon-free, and readable by a junior. Flags bugs, risks, and nits separately. Asks the code owner when intent is unclear rather than assuming. |
Two hooks run automatically:
| Event | Hook | Effect |
|---|---|---|
SessionStart | caveman-activate.js | Injects caveman rules as session context so every session starts in caveman mode |
UserPromptSubmit | caveman-mode-tracker.js | Tracks active mode to ~/.claude/.caveman-active — used by statusline scripts to show the current caveman level |
npx claudepluginhub preludium/claude-kit --plugin preludium-skillsHarness-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
Feature development with code-architect/explorer/reviewer agents, CLAUDE.md audit and session learnings, and Agent Skills creation with eval benchmarking from Anthropic.
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.