Claude Code core developer Thariq announces departure from Anthropic
Claude Code core developer Thariq announced his departure after working at Anthropic for over a year, saying he witnessed Claude Code scale far beyond what he had imagined. Claude Code is an autonomous coding agent that resides in the CLI, able to pull branches, read codebases, locate bugs, run tests, and submit git commits on its own. Thariq said he will first take some time off, then move on to new things.
View original postI’ve left Anthropic after more than a year of working on Claude Code and seeing it scale larger than I ever imagined. I have really loved working at Ant and believe deeply in the mission, and most of all, the idea of a race to the top. The idea that if we work hard and responsibly, we can raise the bar and inspire others to great outcomes. And if we all push each other to be better, things can be radically good. I will carry the spirit of this with me forever. My colleagues at Anthropic are the most talented, thoughtful, competent, and earnest people I have ever met. I am glad they are at the wheel and trust them deeply. To the community on X, thank you for everything. You’ve made this the most fun (and sometimes stressful) job I’ve ever had. For me, it is time for a bit of a break before diving into something new. If you want more excellent Claude Code content, you should follow @addyosmani , @lydiahallie, @_catwu, @ClaudeDevs and of course @bcherny who started it all.
View original
I’ve left Anthropic after more than a year of working on Claude Code and seeing it scale larger than I ever imagined. I have really loved working at Ant and believe deeply in the mission, and most of all, the idea of a race to the top. The idea that if we work hard and responsibly, we can raise the bar and inspire others to great outcomes. And if we all push each other to be better, things can be radically good. I will carry the spirit of this with me forever. My colleagues at Anthropic are the most talented, thoughtful, competent, and earnest people I have ever met. I am glad they are at the wheel and trust them deeply. To the community on X, thank you for everything. You’ve made this the most fun (and sometimes stressful) job I’ve ever had. For me, it is time for a bit of a break before diving into something new. If you want more excellent Claude Code content, you should follow @addyosmani , @lydiahallie, @_catwu, @ClaudeDevs and of course @bcherny who started it all.
Body · AI translation
Damn, I'm completely shocked!! Claude Code, the tool that pulled countless programmers out of overtime, today its most core operator behind it, Thariq, announced his departure 😭 Thariq worked at Anthropic for over a year, personally raising this super agent that runs in the terminal; these few sentences in his farewell letter directly reveal the underlying secret of the next generation of tool explosion. Brothers, if you've recently used Claude's official command-line terminal tool, you've most likely had your worldview shattered by its autonomous refactoring capabilities. And the core member who personally raised this nuclear-grade tool from lines of code, today officially announced leaving Anthropic 🤯
About Thariq (@trq212), many non-developers may not be familiar with this name yet.
But in the entire AI coding and terminal agent circle, he is absolutely one of the most hardcore behind-the-scenes drivers of the past year.
He stayed at Anthropic for over a year, and his core achievement was as a core developer, leading and participating in the development of Claude Code, which has just swept the entire internet.
Many people think the endgame of large models writing code is the web chat box, or editor autocomplete like Cursor.
But Thariq and his team directly stuffed Claude into the darkest, lowest-level terminal of countless programmers.
It doesn't require you to copy and paste back and forth between the web page and the editor; instead, it directly acts as a fully autonomous engineer residing in the CLI: give it a requirement, and it pulls branches itself, reads your entire codebase of hundreds of thousands of lines, locates bugs, runs test cases, and even fixes them itself and directly submits a git commit.
Thariq wrote a particularly thought-provoking line in his farewell letter, which moved me deeply: witnessed it scale far beyond what I imagined.
What may have originally been a small toy written by engineers internally just to be lazy quickly became, in a very short time, the standard workflow reshaping the daily habits of tens of millions of developers worldwide.
This is exactly the most terrifying growth logic of this generation of top AI products: it no longer needs a huge sales team knocking on doors to sell; as long as the tool solves an engineer's bone-deep tedious debugging pain once in the terminal, word of mouth will complete fission in the developer community at the speed of light.
I think what he takes away is not only this year's achievements, but more critically the product philosophy he leaves behind: Race to the top.
Using truly responsible hardcore engineering to pull industry standards extremely high, forcing peers around the world to come up with something greater to compete.
Truly penetrating tools in the large model era have never been about redrawing the tables of the old world, but rather, like Claude Code, directly drilling into the most hardcore productivity scenarios and uprooting the traditional toolchain.
Salute to the behind-the-scenes hero who built a god-tier productivity foundation. What's most worth watching next is, after such a top-tier full-stack engineering brain leaves, where the next fire will be opened from. https://t.co/wedFErrDTw