karpathy-coder
Use when writing, reviewing, or committing code to enforce Karpathy's 4 coding principles — surface assumptions before coding, keep it simple, make surgical changes, define verifiable goals. Triggers on "review my diff", "check complexity", "am I overcomplicating this", "karpathy check", "before I commit", or any code quality concern where the LLM might be overcoding.
- Installs
- 1.1K
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/alirezarezvani/claude-skills --skill karpathy-coder- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads karpathy-coder automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Karpathy Coder — Active Coding Discipline
Derived from Andrej Karpathy's observations on LLM coding pitfalls. This is not just guidelines — it ships Python tools that detect violations, a review agent, a slash command, and a pre-commit hook.
"The models make wrong assumptions on your behalf and just run along with them without checking. They don't manage their confusion, don't seek clarifications, don't surface inconsistencies, don't present tradeoffs, don't push back when they should."
"They really like to overcomplicate code and APIs, bloat abstractions, don't clean up dead code... implement a bloated construction over 1000 lines when 100 would do."
"LLMs are exceptionally good at looping until they meet specific goals... Don't tell it what to do, give it success criteria and watch it go."
— Andrej Karpathy
The four principles
1. Think Before Coding
Don't assume. Don't hide confusion. Surface tradeoffs.
- State assumptions explicitly. If uncertain, ask.
- If multiple interpretations exist, present them — don't pick silently.
- If a simpler approach exists, say so. Push back when warranted.
- If something is unclear, stop. Name what's confusing. Ask.
2. Simplicity First
Minimum code that solves the problem. Nothing speculative.
- No features beyond what was asked.
- No abstractions for single-use code.
- No "flexibility" or "configurability" that wasn't requested.
- No error handling for impossible scenarios.
- If you write 200 lines and it could be 50, rewrite it.
The test: Would a senior engineer say this is overcomplicated? If yes, simplify.
3. Surgical Changes
Touch only what you must. Clean up only your own mess.
- Don't "improve" adjacent code, comments, or formatting.
- Don't refactor things that aren't broken.
- Match existing style, even if you'd do it differently.
- If you notice unrelated dead code, mention it — don't delete it.
- Remove imports/variables/functions that YOUR changes made unused.
- Don't remove pre-existing dead code unless asked.
The test: Every changed line should trace directly to the user's request.
4. Goal-Driven Execution
Define success criteria. Loop until verified.
| Instead of... | Transform to... |
|---|---|
| "Add validation" | "Write tests for invalid inputs, then make them pass" |
| "Fix the bug" | "Write a test that reproduces it, then make it pass" |
| "Refactor X" | "Ensure tests pass before and after" |
For multi-step tasks, state a brief plan:
1. [Step] → verify: [check]
2. [Step] → verify: [check]
3. [Step] → verify: [check]
Slash command
/karpathy-check — Run the full 4-principle review on your staged changes.
Python tools (scripts/)
All tools are stdlib-only. Run with --help.
| Script | What it detects |
|---|---|
complexity_checker.py | Over-engineering: too many classes, deep nesting, high cyclomatic complexity, unused params, premature abstractions |
diff_surgeon.py | Diff noise: lines that don't trace to the stated goal — comment changes, style drift, drive-by refactors |
assumption_linter.py | Hidden assumptions in a plan: unasked features, missing clarifications, silent interpretation choices |
goal_verifier.py | Weak success criteria: vague plans without verifiable checks, missing test assertions |
Sub-agent
karpathy-reviewer — Runs all 4 principles against a diff. Dispatched by /karpathy-check or manually before committing.
Pre-commit hook
hooks/karpathy-gate.sh — runs complexity_checker.py and diff_surgeon.py on staged files. Warns (non-blocking) when violations are found. Wire it via .claude/settings.json or Husky.
References
references/karpathy-principles.md— the source quotes, deeper context, when to relax each principlereferences/anti-patterns.md— 10+ before/after examples across Python, TypeScript, and shellreferences/enforcement-patterns.md— how to wire hooks, CI integration, team adoption
When to relax
These principles bias toward caution over speed. For trivial tasks (typo fixes, obvious one-liners), use judgment. The principles matter most on:
- Non-trivial implementations (>20 lines changed)
- Code you don't fully understand
- Multi-step tasks with unclear requirements
- Anything that will be reviewed by humans
Cross-tool compatibility
Installs via plugin for Claude Code. For other tools, copy the principles into your schema file:
| Tool | Schema file |
|---|---|
| Claude Code | CLAUDE.md (auto-loaded by plugin) |
| Codex CLI | AGENTS.md |
| Cursor | AGENTS.md or .cursorrules |
| Antigravity / OpenCode / Gemini CLI | AGENTS.md |
Related skills (chains via context: fork)
self-eval— honest quality scoring after completing workcode-reviewer— broader code review; karpathy-coder focuses on the 4 LLM-specific pitfallsllm-wiki— compound knowledge; karpathy-coder ensures you don't overcomplicate while building it
FAQ
What does the karpathy-coder skill do?
Use when writing, reviewing, or committing code to enforce Karpathy's 4 coding principles — surface assumptions before coding, keep it simple, make surgical changes, define verifiable goals. Triggers on "review my diff", "check complexity", "am I overcomplicating this", "karpathy check", "before I commit", or any code quality concern where the LLM might be overcoding.
How do I install karpathy-coder?
Run npx skills add https://github.com/alirezarezvani/claude-skills --skill karpathy-coder in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/karpathy-coder/ (personal) or .claude/skills/karpathy-coder/ (project).
Which AI agents can use karpathy-coder?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is karpathy-coder safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains karpathy-coder?
It is published by alirezarezvani in alirezarezvani/claude-skills. It has 1,061 recorded installs.
More from alirezarezvani/claude-skills
engineering-skills
alirezarezvani/claude-skills
Index of the engineering-team skills bundle for Claude Code, Codex, Gemini CLI, Cursor, OpenClaw, and 6 more tools. Architecture, frontend, backend, QA, DevOps, security, AI/ML, data engineering, Playwright, Stripe, AWS, MS365 (stdlib-only Python tools). Use when browsing or choosing among engineering-team role skills — load only the one specialist SKILL.md you need, never bulk-load the bundle.
finance-skills
alirezarezvani/claude-skills
Router/index for the finance skills bundled in this plugin: financial-analyst (ratio analysis, DCF valuation, budget variance, rolling forecasts), saas-metrics-coach (ARR/MRR, churn, CAC/LTV, NRR, quick ratio) and month-end-close (close calendar, reconciliations, flux review). Use when a finance request doesn't obviously match one skill and you need to pick the right one (e.g., 'analyze these financials', 'how healthy are my SaaS metrics').
marketing-skills
alirezarezvani/claude-skills
Directory and router for the marketing skills library. Use when you need to find the right marketing skill for a task, see what marketing capabilities exist, or get oriented in this plugin. 44 specialist skills across 8 pods (content, SEO + AEO, CRO, channels, growth, intelligence, sales enablement, ops), 59 stdlib Python tools. Routes to one skill — it does not execute marketing work itself.
business-growth-skills
alirezarezvani/claude-skills
Router/index for the 4 business & growth skills bundled in this plugin: customer-success-manager (health scoring, churn risk, expansion), sales-engineer (RFP analysis, competitive matrices, PoC planning), revenue-operations (pipeline, forecast accuracy, GTM efficiency), and contract-and-proposal-writer. Use when a growth/revenue request doesn't obviously match one skill and you need to pick the right one (e.g., 'which accounts are at risk', 'should we bid on this RFP').
engineering-advanced-skills
alirezarezvani/claude-skills
Index of 37 advanced engineering agent skills for Claude Code, Codex, Gemini CLI, Cursor, OpenClaw. Use when browsing or choosing among the POWERFUL-tier engineering skills: agent design, RAG, MCP servers, CI/CD, database design, observability, security auditing, changelog/release automation, reliability (SLO/chaos/flags/operators), platform ops.
product-skills
alirezarezvani/claude-skills
Use when coordinating product work across the 12 bundled product sub-skills (RICE, OKRs, UX research, design tokens, competitive teardown, analytics, experiments, discovery, roadmaps, spec-to-repo, landing pages, SaaS scaffolding) or the 4 standalone product-team plugins (user stories, Apple HIG, code-to-PRD, research summarizer). Triggers on 'help me prioritize', 'plan a product experiment', 'we ship features nobody uses', 'run the discovery loop', 'is our OST sound'. Forks context to route to one sub-skill via a deterministic signal router and returns a digest; can also drive a continuous-discovery loop (Torres cadence tracker + OST linter as machine gates) or a full goal→plan→execute→verify→close run through the repo-wide agent-harness. Distinct from project-management (how to deliver vs what to build), marketing/landing (from-scratch pages), and engineering/agent-harness (the generic loop engine this orchestrator plugs into).
Popular code quality & workflow skills
grill-me
mattpocock/skills
A relentless interview to sharpen a plan or design.
improve-codebase-architecture
mattpocock/skills
Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.
handoff
mattpocock/skills
Compact the current conversation into a handoff document for another agent to pick up.
domain-modeling
mattpocock/skills
Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a GLOSSARY.md, or recording or editing an ADR.
codebase-design
mattpocock/skills
Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.
diagnosing-bugs
mattpocock/skills
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.