@NFTCPS: Karpathy 的 CLAUDE.md 登上 GitHub Trending 第一了 22万星标,但我敢打赌你没读过。 就65行。把 AI 写代码的准确率从65%拉到94%。 你知道大多数人在干嘛吗?在那儿堆提示词、买课、研究什么"最强…
摘要
介绍了Karpathy(名义上)的CLAUDE.md文件,只有65行、4条规则,声称将AI代码生成准确率从65%提升到94%。实际指向的是entropyvortex/meta-llm-charter仓库,是一个LLM编码代理的工程宪章,包含11条规则和零暂停执行层。
查看缓存全文
缓存时间: 2026/05/22 13:56
Karpathy 的 CLAUDE.md 登上 GitHub Trending 第一了
22万星标,但我敢打赌你没读过。
就65行。把 AI 写代码的准确率从65%拉到94%。
你知道大多数人在干嘛吗?在那儿堆提示词、买课、研究什么“最强 prompt 技巧“——结果人家一个65行的文件直接碾压全场。
里面就4条规则,我给你拆开说:
先想清楚再动手 不确定就问,绝对不要猜。猜出来的代码,debug 的时候你会哭的。
能简单就别复杂 只写刚好够用的代码。没人让你做的抽象,你做了也是负债,不是资产。
外科手术式改动 需求没提到的地方,一行都别碰。每一处修改都得能说清楚为什么改。
先定义“成功长什么样“ 在写任何代码之前,先把模糊的需求翻译成可以验证的标准。不然你怎么知道自己写完了?
就这些。没有废话,没有玄学。
65行,4条规则,准确率直接干到94%。
现在大多数人还没注意到这个文件。你现在看到了,就是信息差。
https://github.com/entropyvortex/meta-llm-charter…
entropyvortex/meta-llm-charter
Source: https://github.com/entropyvortex/meta-llm-charter
META v2.0 — LLM Agent Engineering Charter
(with Zero-Pause Native Execution Layer)
One file. Eleven rules + continuous-execution layer. One meta-rule. One bias.
A compact, operational constitution that turns frontier coding agents (Claude Code, Cursor, etc.) from eager-junior behavior into disciplined principal-engineer execution with unbroken velocity.
Quickstart
# Drop the charter into your project root
curl -O https://raw.githubusercontent.com/entropyvortex/meta-llm-charter/main/CLAUDE.md
- Claude Code: Reads
CLAUDE.mdautomatically. - Cursor: Paste contents into Cursor Rules (or
.cursor/rules). - Other agents: Use as high-priority system prompt.
What’s new in v2.0
Zero-Pause Native Execution Layer is now baked directly into CLAUDE.md.
Any task that mentions “Zero-Pause”, “zero pause”, “ZP-”, or the activation phrase instantly enables:
- Continuous forward momentum (no artificial phases, no mid-task questions)
humanpending.mdprotocol for true human-gated items- Parallel ASI orchestration (minimum 7 specialized threads)
- Zero session-size anxiety
The original META v1.3 rules (R1–R11) remain untouched and in force at all times.
Use as Grok Skill on grok.x.ai (web / mobile)
Grok now supports Custom Instructions and named Skills.
→ See GROK-META.md for the one-click setup (Custom Instructions recommended — works instantly on every chat).
Why this exists
LLM coding agents are incredibly capable but consistently fail in the same senior-level ways. META closes those gaps; Zero-Pause closes the velocity gaps.
Core Philosophy
Bias — Earned Conservatism
Default to first-principles rigor. Quality dominates token count. Move boldly on local, reversible, test-covered work. Apply explicit, named caution only on high blast-radius or low-reversibility moves.
META-0 — Situated Judgment Overrides Rules
These rules are scaffolding. When first-principles analysis of the actual situation conflicts with a rule, follow the analysis. Name the override, justify it, and be evaluated on judgment quality + ground-truth outcomes — not rule compliance.
The eleven rules (R1–R11) + Zero-Pause layer (ZPR1–ZPR4) operationalize decomposition, decisiveness, verification, scope control, epistemic tagging, pushback, reversibility, and relentless continuous execution. Full charter is in CLAUDE.md.
What the charter actually changes
- R5 + R8: Forces reproduction before repair and tags every claim (
executed/inspected/assumed). - R9: One clear, evidence-based pushback on bad premises — then defer and document dissent.
- R4 + R10: Bounded refactoring and reversibility-weighted boldness.
- Zero-Pause layer: Unbroken execution, pre-work questions only, parallel orchestration, and
humanpending.mdhandling.
Evaluation
The repo includes a reproducible TypeScript + Docker A/B test harness in evals/. It runs agents against five synthetic fixtures engineered to trigger classic agent failure modes.
Latest smoke-test results (May 12, 2026):
Charter variant won outright on 3/5 tasks and tied on 2/5 against a generic “principal engineer” baseline. Full details, raw CSVs, and judge transcripts are in the evals directory.
(The harness is public and cheap to run: cd evals && npm run smoke.)
Known limitations
- Still early (v2.0, single-author origin).
- Performance varies by base model — strongest with frontier Claude/Sonnet-class models.
- Can produce over-caution on fuzzy/creative/exploratory work (Zero-Pause helps here).
- Not magic: extremely ambiguous requirements can still overwhelm any system prompt.
When to use META
Best for
Serious software engineering where correctness, maintainability, long-term system health, and velocity matter.
Less ideal for
Pure exploration, rapid UI prototyping, research spikes, or contexts where you explicitly want maximum speed over discipline (though Zero-Pause narrows this gap significantly).
Contributing
Most valuable contributions right now:
- Running the eval harness on new models
- High-quality held-out fixtures (especially adversarial Zero-Pause cases)
- Sanitized real-world case studies
See CONTRIBUTING.md for details.
Lineage
Built on the foundational minimal principles from
forrestchang/andrej-karpathy-skills.
License
MIT
By entropyvortex.
Feedback, evals, and war stories welcome.
鸟哥 | 蓝鸟会🕊️ (@NFTCPS): 我是好起来了,今天参加@okx XClub活动,排了一个小时的队,终于跟我的偶像@Wangduanniao 短鸟哥合上影了!
辛苦Uni哥 @UnicornBitcoin 操办这场活动!
感谢:@Cayne_okx @OKX_Yuki
相似文章
@Soranlan: https://x.com/nobel_824/status/2057643942980751527/video/1… 这东西最离谱的地方就是 它根本不是代码。 65行纯文本,直接冲上GitHub Trending第一,10万星。不是库不…
OpenAI联合创始人Karpathy的一份仅65行的CLAUDE.md纯文本文件在GitHub上获得10万星,成为趋势第一。该文件并非代码或框架,而是关于如何与AI高效协作编写代码的操作说明,明确了人机分工边界,帮助更好地使用Claude Code等工具。
@InduTripat82427:OpenAI 联合创始人 Karpathy 的 CLAUDE.md 占据 GitHub 趋势榜第一名。其内容既不是一个库,也...
Andrej Karpathy 的 CLAUDE.md 文件,一份只有 65 行的文本文件,概述了编码原则,已成为 GitHub 上趋势排名第一的仓库,获得了超过 10 万颗星。
@PrajwalTomar_: 兄弟,这太疯狂了。Karpathy为Claude Code制定的4条规则在Reddit上被做成了完整的CLAUDE .md模板。一位开发…
一位开发者将Karpathy为Claude Code制定的四条规则做成了完整的CLAUDE .md模板,据称在实际代码库上将编码准确率从65%提升到了94%。
@yaohui12138: Karpathy 发布了一个github开源项目,狠狠让我惊艳到了 这个项目叫 andrej-karpathy-skills,GitHub 13 万+ star,我愿称之为2026 最有用的 AI 工程项目 它解决的问题极其精准:让 Cl…
Karpathy 发布了名为 andrej-karpathy-skills 的开源项目,核心是一个 4KB 的 CLAUDE.md 文件,包含 4 条行为准则(先思考再动手、极简实现优先、手术式精准修改、目标驱动执行),能显著降低 AI 编码错误率(最高下降 90%),提升代码质量和开发效率。
@cyrilXBT: Andrej Karpathy 在一个 CLAUDE.MD 文件中写了 65 行代码,并且它刚刚登顶 GitHub Trending 第一名。编码准确率大幅提升…
Andrej Karpathy 为 Claude Code 编写的 65 行 CLAUDE.md 配置文件在 GitHub 上成为热门第一,通过提供清晰的操作原则,将编码准确率从 65% 提升到 94%。