[flagged]
Prompt injection from github issue? This is funny but actually may be true.
Claude Code's source code has been leaked via a map file in their NPM registry
421–430 of 1001 posts
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#422Earlier quoted context omitted.
They stole everything and now they want to close the gates behind them. "I got the loot, Steve!" I feel like the distillation stuff will end up in court if they try to sue an American company about it. We'll see what a judge says.
You're perfectly free to scrape the web yourself and train your own model. You're not free to let Anthropic do that work for you, because they don't want you to, because it cost them a lot of time and money and secret sauce presumably filtering it for quality and other stuff. Stole? Courts have ruled it's transformative, and it very obviously is. AI doomerism is exhausting, and I don't even use AI that much, it's jus…
Guess who else spent a lot of time and money and secret sauce?
Do you hear the words coming out of your mouth?
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#423They have an interesting regex for detecting negative sentiment in users prompt which is then logged (explicit content): https://github.com/chatgptprojects/claude-code/blob/642c7f94... I guess these words are to be avoided...
I'm clearly way too polite to Claude. Also: // Match "continue" only if it's the entire prompt if (lowerInput === 'continue') { return true } When it runs into an error, I sometimes tell it "Continue", but sometimes I give it some extra information. Or I put a period behind it. That clearly doesn't give the same behaviour.
I always type "please continue". I guess being polite is not a good idea.
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#424Earlier quoted context omitted.
Think about your question, depending on the tool, Rust might not be needed, is high level memory performance and safety needed in a coding agent ? Probably not. It's high speed iteration of release ? Might be needed, Interpreted or JIT compiled ? might be needed. Without knowing all the requirements its just your workspace preference making your decision and not objectively the right tool for the job.
I have a 16GB RAM laptop. It's a beast I bought in 2022. It's all I need for my work. RAM on this machine can't be upgraded. No issue when running a few Codex instances. Claude: forget it. That's why something like Rust makes a lot of sense. Even more now, as RAM prices are becoming a concern.
> Claude: forget it.
I don't know what else you're doing but the footprint of Claude is minor.
Anyway my point still stands, you're looking at it as if they are competing languages and one is better at all things. That just not how things work.
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#425They have an interesting regex for detecting negative sentiment in users prompt which is then logged (explicit content): https://github.com/chatgptprojects/claude-code/blob/642c7f94... I guess these words are to be avoided...
An LLM company using regexes for sentiment analysis? That's like a truck company using horses to transport parts. Weird choice.
The amount of trust and safety work that depends on google translate and the humble regex, beggars the imagination.
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#426Earlier quoted context omitted.
You'd be playing cat and mouse like yt-dlp, but there's probably more value to this code than just a temporary way to milk claude subscriptions.
If you're using a claude subscription you'd just use claude code. The real value here will be in using other cheap models with the cc harness.
Lol what? There is no value. OpenCode and Pi and more exist. Arguably Claude Code is the worst client on the market. People use Claude Code not because it's some amazing software. It's to access Opus at a discounted rate.
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#427Earlier quoted context omitted.
They stole everything and now they want to close the gates behind them. "I got the loot, Steve!" I feel like the distillation stuff will end up in court if they try to sue an American company about it. We'll see what a judge says.
You're perfectly free to scrape the web yourself and train your own model. You're not free to let Anthropic do that work for you, because they don't want you to, because it cost them a lot of time and money and secret sauce presumably filtering it for quality and other stuff. Stole? Courts have ruled it's transformative, and it very obviously is. AI doomerism is exhausting, and I don't even use AI that much, it's jus…
Let's talk ethics, not law. Why is it okay for these companies to pirate books and scrape the entire web and offer synthesized summaries of all of it, lowering traffic and revenue for countless websites and professions of experts, but it is not okay for others to try to do the same to an AI model?
Is the work of others less valid than the work of a model?
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#428Who cares? It's Javascript, if anyone were even remotely motivated deobfuscation of their "closed source" code is trivial. It's silly that they aren't just doing this open source in the first place.
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#429src/cli/print.ts This is the single worst function in the codebase by every metric: - 3,167 lines long (the file itself is 5,594 lines) - 12 levels of nesting at its deepest - ~486 branch points of cyclomatic complexity - 12 parameters + an options object with 16 sub-properties - Defines 21 inner functions and closures - Handles: agent run loop, SIGINT, rate-limits, AWS auth, MCP lifecycle, plugin install/refresh, wo…
"That's Larry; he does most of the work around here."
Re: Claude Code's source code has been leaked via a map file in their NPM registry
#430ANTI_DISTILLATION_CC This is Anthropic's anti-distillation defence baked into Claude Code. When enabled, it injects anti_distillation: ['fake_tools'] into every API request, which causes the server to silently slip decoy tool definitions into the model's system prompt. The goal: if someone is scraping Claude Code's API traffic to train a competing model, the poisoned training data makes that distillation attempt less…
Why would this be in the client code though?