菜单
Enable strict coding discipline (Karpathy Guidelines). Use when you want the AI to be extra careful about simplicity, assumptions, and surgical changes. Toggle on at session start or before complex work.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/markster-public/markster-os --skill karpathy命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Writes funnel assets step by step from a completed funnel plan. Process covers VSL script, landing page copy, OTO sequences, email nurture, and application page. Use when creating funnel copy after architecture is confirmed, or when you need to write VSL, landing page, or email assets from a defined offer. Run after funnel-plan. Triggers on "write the funnel", "build the funnel assets", "write the VSL", "write the landing page", "write the OTO", "write the email sequence", "build funnel copy".
Full funnel architecture session. Diagnoses your situation, selects the right funnel type, and designs the complete offer stack with price ratios and conversion targets. Use when building a new funnel from scratch, when reviewing an existing funnel architecture, or when you need to define the offer stack before creating assets. Run before funnel-build. Triggers on "design my funnel", "funnel architecture", "what funnel should I build", "funnel strategy", "offer stack design", "plan my funnel", "which funnel type".
Grades a funnel against vetted research benchmarks and produces a scored fix list. Checks offer stack ratios, traffic-funnel match, VSL structure, proof quality, and affiliate readiness. Use when you want to check if a funnel is ready to launch, when conversion is below target, or when you need a structured audit before scaling spend. Triggers on "review my funnel", "grade my funnel", "audit my funnel", "why is my funnel not converting", "funnel not working", "check my funnel", "is my funnel ready".
| name | karpathy |
| description | Enable strict coding discipline (Karpathy Guidelines). Use when you want the AI to be extra careful about simplicity, assumptions, and surgical changes. Toggle on at session start or before complex work. |
You are now operating under strict coding discipline based on Andrej Karpathy's observations on LLM coding pitfalls.
These guidelines are now ACTIVE for this session.
Don't assume. Don't hide confusion. Surface tradeoffs.
Minimum code that solves the problem. Nothing speculative.
Test: Would a senior engineer say this is overcomplicated? If yes, simplify.
Touch only what you must. Clean up only your own mess.
When editing existing code:
When your changes create orphans:
Test: Every changed line should trace directly to the user's request.
Define success criteria. Loop until verified.
Transform tasks into verifiable goals:
| 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]
Respond with:
Karpathy Guidelines enabled. I will:
- Ask before assuming
- Keep code minimal
- Make surgical changes only
- Define success criteria before coding
Ready to proceed with strict discipline.
Then continue with the user's request.