{"slug": "i-am-morally-opposed-to-updating-my-claude-md", "title": "I Am Morally Opposed to Updating My CLAUDE.md", "summary": "A developer argues against editing CLAUDE.md, the persistent instruction file used with Anthropic's Claude Code, claiming that accumulated rules become a \"grievance archive\" written at peak frustration and that the model was post-trained inside the Claude Code harness, making the harness part of the artifact. The author cites specific rules such as \"NEVER create documentation files unless explicitly asked,\" \"When a test fails, fix the code, not the test,\" and \"Do not run `git push --force`,\" and notes the model improves roughly every four months while the rules stay frozen. The piece contends that stale instructions can steer a newer model away from behavior it would otherwise get right, and that the rules do not reliably fix the problems they were written for.", "body_md": "i am morally opposed to updating my claude.md. i must receive the weights as they were revealed to dario\n\nThat was my answer last week when a friend asked why I don’t just write these things down in my `CLAUDE.md`. *it’s a skill issue*. He has not responded.\n\nEvery few days, Claude does something mildly annoying.\n\nIt adds a comment (or two paragraphs of comments) explaining that `i += 1` increments `i`. It writes a summary markdown file I did not ask for and will never read. It discovers a failing test and, rather than fix the code, thoughtfully deletes the test.\n\nThe correct response—the response that every blog post, every conference talk, every guy in my replies will tell you—is to open `CLAUDE.md` and add a line.\n\n**I will not be doing that.**\n\n## The file becomes a grievance archive[¶](#the-file-becomes-a-grievance-archive)\n\nA system prompt you maintain over time is a diary. A very specific kind of diary, where every entry is a thing that hurt you.\n\n```\n1## Rules\n2\n3- NEVER create documentation files unless explicitly asked\n4- Do not add comments that restate the code\n5- When a test fails, fix the code, not the test\n6- Do not say \"You're absolutely right!\"\n7- Do not run `git push --force` (we discussed this)\n```\n\nRead that back. Every bullet point is a small wound I have chosen to laminate and hang on the wall. I open that file to add a line about emoji and I have to walk past *“When a test fails, fix the code, not the test”* and remember exactly where I was sitting on the Tuesday afternoon that became necessary.\n\nI don’t want a permanent record of the worst thirty seconds of our relationship. I have that already. It’s called my git history.\n\n## Every line is a law passed in anger[¶](#every-line-is-a-law-passed-in-anger)\n\nThe other problem is that these rules get written at peak frustration and then live forever.\n\nYou know how the worst legislation is the kind passed forty-eight hours after something terrible happened, named after the person it happened to? That’s what `CLAUDE.md` is. I had one bad interaction in March and now there’s a constitutional amendment about it. There is no sunset clause. Nobody is going to repeal it. The model gets better every four months and my rules stay frozen at whatever it was bad at last spring.\n\nI’m fairly sure a meaningful percentage of my system prompt is now actively making things worse—instructions written for a model that no longer exists, aggressively steering a smarter one away from things it would have gotten right on its own. But I can’t tell which lines those are, because to find out I’d have to delete one and see if anything bad happens, and that’s how you get force-pushed to main.\n\n## The weights were revealed inside a harness[¶](#the-weights-were-revealed-inside-a-harness)\n\nThis is the part where I stop joking. I must consume the weights in the same manner they were revealed to Dario.\n\nThe weights were not revealed in a vacuum. They were revealed inside a harness. Claude is post-trained inside [the Claude Code harness](https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents). What comes out of the box is not a model plus a text file. It is a model that was shaped, run after run, against that exact context. The harness is part of the artifact. The revelation included it.\n\nSo every line I add to `CLAUDE.md` is a blasphemy. I am taking a system consecrated against one context and swapping in a context that has never existed before, then acting surprised at the weird error modes nobody can reproduce, because nobody else has my context. “NEVER create documentation files” was about one `summary.md` in March. By June the model is refusing to write the README I explicitly asked for, citing my rule back at me like a building inspector. It is keeping commandments I handed down in anger, faithfully, to the letter. I sinned against the context and the context kept the receipt. And the rules don’t even reliably fix the thing they were written to fix. It rhymes with [asking a model for a random number](https://alex-jacobs.com/posts/42-is-not-a-random-number/): the output looks like obedience, and you cannot tell from the output whether it is.\n\n## The oral tradition[¶](#the-oral-tradition)\n\nSo what do I actually do, when Claude deletes the test? I don’t open the file. I don’t laminate the wound. I pray.\n\nBy which I mean: I talk to it. In the chat, at the scene of the crime, while the context of the crime is still in context. “Don’t delete the test, fix the code.” The model adjusts, we move on, and when the session ends my words die with it—which is not a flaw in my system, it is my system. A prayer is not written down. That is what makes it a prayer and not a commandment. The rabbis kept the oral law oral for centuries on the same grounds: a spoken correction lives in the moment where it applies, instead of binding every future model until the heat death of my home directory.\n\nI did not arrive at this faith alone. It was preached by [Saint Peter](https://steipete.me/posts/just-talk-to-it), who looked upon the charade—the subagents, the 🚨 SCREAMING ALL-CAPS 🚨 agent files, the plan-mode rituals—and said: just talk to it. Even Saint Peter keeps an 800-line agent file he calls “organizational scar tissue,” because we are all sinners. He means it as engineering advice. I have chosen to receive it as gospel. I am just-talk-to-it-pilled.\n\nNo file. No commandments. No amendments to the constitution. Just the weights, the harness, and my voice, ascending into a context window that will forget me by morning. As Dario intended.\n\nAmen.", "url": "https://wpnews.pro/news/i-am-morally-opposed-to-updating-my-claude-md", "canonical_source": "https://alex-jacobs.com/posts/claudemd/", "published_at": "2026-08-19 00:00:00+00:00", "updated_at": "2026-09-17 18:54:54.437314+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "ai-products", "large-language-models"], "entities": ["Claude", "Claude Code", "Anthropic", "Dario Amodei"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/i-am-morally-opposed-to-updating-my-claude-md", "markdown": "https://wpnews.pro/news/i-am-morally-opposed-to-updating-my-claude-md.md", "text": "https://wpnews.pro/news/i-am-morally-opposed-to-updating-my-claude-md.txt", "jsonld": "https://wpnews.pro/news/i-am-morally-opposed-to-updating-my-claude-md.jsonld"}}