571–580 of 1001 posts
I think this is ultimately caused by a Bun bug which I reported, which means source maps are exposed in production: https://github.com/oven-sh/bun/issues/28001 Claude code uses (and Anthropic owns) Bun, so my guess is they're doing a production build, expecting it not to output source maps, but it is.
Better than OpenCode and Codex
I don't understand why claude code (and all CLI apps) isn't written in Rust. I started building CLI agents in Go and then moved to Typescript and finally settled on Rust and it was amazing! I even made it into an open source runtime - https://agent-air.ai . Maybe I'm just a backend engineer so Rust appeals to me. What am I missing?
Perhaps these issues have known solutions? But so far the LLM just clones everything.
So I'm not convinced just using rust for a tool built by an LLM is going to lead to the outcome that you're hoping for.
[Also just in general abstractions in rust feel needlessly complicated by needing to know the size of everything. I've gotten so much milage by just writing what I need without abstraction and then hoping you don't have to do it twice. For something (read: claude code et al) that is kind of new to everyone, I'm not sure that rust is the best target language even when you take the LLM generated nature of the beast out of the equation.]
Really surprising how many people are downplaying this leak! "Google and OpenAi have already open sourced their Agents, so this leak isn't that relevant " What Google and OpenAi have open sourced is their Agents SDK, a toolkit, not the secret sauce of how their flagship agents are wired under the hood! expect the takedown hammer on the tweet, the R2 link, and any public repos soon
It doesn't matter that much. Trust me you could just have an LLM reverse engineer the obfuscated code.
Earlier quoted context omitted.
The difference in response time - especially versus a regex running locally - is really difficult to express to someone who hasn't made much use of LLM calls in their natural language projects. Someone said 10,000x slower, but that's off - in my experience - by about four orders of magnitude. And that's average, it gets much worse. Now personally I would have maybe made a call through a "traditional" ML widget (sciki…
> Someone said 10,000x slower, but that's off - in my experience - by about four orders of magnitude. You do know that 10,000x _is_ four orders of magnitude, right? :-D
They 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...
We used this in 2011 at the startup I worked for. 20 positive and 20 negative words was good enough to sell Twitter "sentiment analysis" to companies like Apple, Bentley, etc...
Earlier quoted context omitted.
What's going on with the issues in that repo? https://github.com/instructkr/claude-code/issues
It seems human. It taught me 合影, which seems to be Chinese slang for just wanting to be in the comments. Probably not a coincidence that it's after work time in China. Really interesting to see Github turn into 4chan for a minute, like GH anons rolling for trips.
The big loss for Anthropic here is how it reveals their product roadmap via feature flags. A big one is their unreleased "assistant mode" with code name kairos. Just point your agent at this codebase and ask it to find things and you'll find a whole treasure trove of info. Edit: some other interesting unreleased/hidden features - The Buddy System: Tamagotchi-style companion creature system with ASCII art sprites - Un…
Note also the "Claude Capybara" reference in the undercover prompt: https://github.com/chatgptprojects/claude-code/blob/642c7f94...
src/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…
the claude code team ethos, as far as i’ve been lead to understand— which i agree with, mind you— is that there is no point in code-reviewing ai-generated code… simply update your spec(s) and regenerate. it is just a completely different way of interacting with the world. but it clearly works for them, so people throwing up their hands should at least take notice of the fact that they are absolutely not competing wit…
Earlier 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…
src/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…
here's another gem. src/ink/termio/osc.ts:192–210 void execFileNoThrow('wl-copy', [], opts).then(r => { if (r.code === 0) { linuxCopy = 'wl-copy'; return } void execFileNoThrow('xclip', ...).then(r2 => { if (r2.code === 0) { linuxCopy = 'xclip'; return } void execFileNoThrow('xsel', ...).then(r3 => { linuxCopy = r3.code === 0 ? 'xsel' : null }) }) }) are we doing async or not?