RT @tautologer: i am currently using the Claude Agent SDK for my web-based agent. there are probably better options for this. what are they?
Retweeted by Thariq [UNOFFICIAL]
@trq212-mir-rt.selfhosted.social
Retweets from @trq212-mirr.selfhosted.social
RT @badlogicgames: what i'm trying to say is: go open weights models. you are dearly needed.
RT @RobertHaisfield: I’m *think* effort level is communicated to the model through the system prompt, so if you change that then it invalidates the cache. Could be wrong though — @trq212 ?
RT @simonw: I hope there are QA testing experts out there who are thinking "finally, we don't need software developers any more!", rolling up their sleeves and building really great software right now
RT @raphaelschaad: The most sacred thing we do at YC isn't Demo Day, in my opinion. It's GOH (Group Office Hours). A bit like a recurring self-help group for high-tech founders. I'm still in touch with most of my W20 peers; bonds for life. Today we took it outdoors:
RT @notch: Oh god, is this how it starts? I'm already coming up with use cases, such as converting all my typescript into javascript. Sure, claude, let's see what you got! Can I still make fun of vibe coders if I become one?
RT @farhunky: Some personal news: I wrapped up my time at @sundayrobotics a few months ago, after an amazing year💛💛 (guess Monday is a fitting day for that announcement!)
RT @AnthropicAI: There’s been a lot of speculation about where we stand on open-weights models. We’ve outlined our views in full here: https://www.anthropic.com/news/position-open-weights-models
Our position on open-weights models
Anthropic CEO Dario Amodei on open-weights models
anthropic.com
RT @Lentils80: Claude Opus 5 made this beautiful painterly world in a single HTML file, everything made procedurally It implemented a realistic grass and wind field simulation so that the millions of grass blades on-screen react to the wind
RT @Marzon_: Thank you @trq212 , as always a banger!
Thariq [UNOFFICIAL]@trq212-mirr.selfhosted.social · 2w agoWe removed ~80% of the Claude Code system prompt for our newest models, this is what we've learned about writing system prompts, skills and Claude.MDs for them. http://x.com/i/article/2080703729385512960
RT @jerhadf: Opus 5 has a better FrontierCode score at medium effort than higher effort - despite increasing performance with effort on other evals. Why? 🧵
RT @kentcdodds: I'm going to make a video about this
Thariq [UNOFFICIAL]@trq212-mirr.selfhosted.social · 2w agoWe removed ~80% of the Claude Code system prompt for our newest models, this is what we've learned about writing system prompts, skills and Claude.MDs for them. http://x.com/i/article/2080703729385512960
RT @TheCodeTherapy: I'm really interested in the proclivity that Opus and Fable have for manipulating files through cat, grep, and sed rather than using the harness tools once the context window starts to fill. Could it be considered emergent behavior in the chase for simplicity?
RT @thdxr: everyone should internalize this all those tricks to make the tool work better aren't going to do much there's way more magic to be found in novel ways to apply the tool to your work
Thariq [UNOFFICIAL]@trq212-mirr.selfhosted.social · 2w agoWe removed ~80% of the Claude Code system prompt for our newest models, this is what we've learned about writing system prompts, skills and Claude.MDs for them. http://x.com/i/article/2080703729385512960
RT @alex_erm: Opus 5, snowboarder test, one shot. On par with Fable, ahead of every other model I've run. No visual defects on the first pass, and the sliding physics feel right.
RT @claudeai: Introducing Claude Opus 5. It's a thoughtful and proactive model that comes close to the frontier intelligence of Fable 5 at half the price.
RT @NotionHQ: Now in beta: Notion as code. Define an entire workspace in TypeScript: teamspaces, databases, custom agents, all of it… then deploy it through the API. Build workspaces with coding agents, version-control your setup in git, and reproduce the same setup anywhere you need it.
RT @simonw: I think loops were a short-lived patch for models that couldn't reliably keep working on long problems until they hit a defined goal Fable and GPT-5.6 (and probably Kimi K3 as well) can just do that out of the box, unassisted
Retweeted by Thariq [UNOFFICIAL]@trq212-mir-rt.selfhosted.social · 2w agoRT @wesmckinn: I think loops are bullshit
RT @claudeai: The Claude Security plugin for Claude Code is now available in beta. Scan your changes for vulnerabilities before you commit, or run a full scan across your codebase, all from your terminal on the Claude inference you already run.
RT @wesmckinn: I think loops are bullshit
RT @notch: Wow, that's a lot of replies with tips and suggestions on how to try vibe coding. I appreciate it! Still probably will be grumpy about it for a while, even if it is good, but who am I to resist the inevitable?
RT @OfficialLoganK: We have started our most ambitious pre-training run yet, for Gemini 4, and are excited by the progress : )
RT @simonw: I got some really useful Claude prompting tips from @_catwu and @trq212 - it's time to stop overloading our prompts with examples and lists of things not to do, Fable works better without those Claude Code's own system prompt recently shrunk by 80%!
RT @simonw: I interviewed @trq212 and @_catwu from the Claude Code team at @aiDotEngineer a couple of weeks ago - the video is now out, so I've published an annotated transcript of our conversation https://simonwillison.net/2026/Jul/21/cat-and-thariq/
A Fireside Chat with Cat and Thariq from the Claude Code team
Earlier this month I hosted a fireside chat session at the AI Engineer World’s Fair with Cat Wu and Thariq Shihipar from Anthropic’s Claude Code team. We talked about Claude …
simonwillison.net
RT @thereturningdev: I missed the memo? do Fable 5 runs on credit only from... now? I have the Max 5x plan
RT @evantana: The fastest way to build is to find the cheapest way to learn. Claude Code's @trq212 talked about this when he came by @southpkcommons.
Retweeted by Thariq [UNOFFICIAL]@trq212-mir-rt.selfhosted.social · 3w agoRT @evantana: AI coding is changing faster than almost anyone can keep up. It's impossible to predict exactly where we'll be a year from now. Claude Code's @trq212 (@southpkcommons alum and 2x founder) came back to SPC to talk about what's on the frontier. (00:00) From Startup Founder to Anthropic
RT @andyfang: Today we're opening up the DoorDash CLI in limited beta. `dd-cli` lets you order DoorDash directly from your agent: search stores, find the best deals, check out, and more. Early access for US/Canadian macOS developers by waitlist. Excited to see what folks build!
RT @GergelyOrosz: I'll be posting a deepdive on this, having talked w sw engineers inside Anthropic's HQ about what they do. The short of it: devs do a lot more+broader work than before; engineers do the design of systems (not AI); coordination v important. More to come in @Pragmatic_Eng
Retweeted by Thariq [UNOFFICIAL]@trq212-mir-rt.selfhosted.social · 3w agoRT @svpino: Why is Anthropic hiring Software Engineers? Why aren't they using agents with loops to do all of this?
RT @svpino: Why is Anthropic hiring Software Engineers? Why aren't they using agents with loops to do all of this?