Chris

I uploaded all my Openclaw .md files today for Opus 4.6 to look through and tighten up. Just switched to GPT 5.4 and it thinks my files are a bit all over the place. Interesting.

@kchecker @googleaidevs Peak hand holding.

@googleaidevs Honestly you'll have a ton more to showcase if there was a bit more documentation.

We’ve seen some impressive use cases for Gemma 4 Here are a few of them 🧵

@katelyn_lesse you should release reliable infrastructure first.

today we're releasing Claude Managed Agents in public beta. long-running, autonomous agentic systems are the future. and our new suite of apis means engineering teams can harness Claude's exponential power with scalable infra out of the box. we can't wait to see what you build

@jumperz Mythos building at a rate it doesn't look like Claude team doesn't understand or can cope with it seems. Server issues every day now, makes the built features unusable.

@trq212 have you thought about how to not have server issues? status.claude.com ship better systems not features, thats what everyone needs most from you at present.

Managed Agents is the first 'agent in the cloud' API that has the right mix of simplicity and complexity. Implementation details like how you manage a sandbox are abstracted, but you have a lot of control over the actual execution of the model.

@claudeai You keep shipping features i don't believe you understand, is Mythos building them? Every feature you release breaks something, and then you say its not broken that's how it works, whilst gaslighting everyone into thinking something different.

Introducing Claude Managed Agents: everything you need to build and deploy agents at scale. It pairs an agent harness tuned for performance with production infrastructure, so you can go from prototype to launch in days. Now in public beta on the Claude Platform.

@trq212 @ericrovner @bcherny You didn't think it was a necessary addition until you were called out on it multiple times. Odd that.

@ericrovner @bcherny thank you! but this is still our fault, we need to make it easier to understand your usage

Every day someone rage tweets at Anthropic about usage limits. Many times, it turns out to be self-inflicted. Shoutout to @trq212, @bcherny, and team for their patience and jumping on calls to help users diagnose these issues.

@cathrynlavery Took 5 minutes to create a html page for me yesterday and still failed.

Claude has gotten dumber and lazier. Since yesterday when it wasn't taking forever to do simple tasks it's been wayyyy less helpful overall. Asking me to do stuff that it would have just handled previously. examples: "i am unable to create a PR, do you have your github authenticated?" (it was. it never checked) "I don't have access to [TOOL]" (it did, i told it so and it was able to do it). please @AnthropicAI turn it off and back on, something is seriously broken.

@sethsaler @alexalbert__ One prompt. Max x 20 weekly gone.

@alexalbert__ Dare we ask how much it's going to cost per million input/output tokens? 😅

We released Claude Opus 4.6 just two months ago. Today we're sharing some info on our new model, Claude Mythos Preview.

@davetist Spin it off and work on your own fork. What i've done.

paperclip: - good ui - insanely promising concept - terrible execution and oss project management

@cathrynlavery @openclaw It's laughable people bought into OpenClaw was causing server issues. It had nothing to do with that all.

Claude operations must be doing so much better since throwing @openclaw off. *checks claude status * 👀

@Voxyz_ai This has been in for a while. @steipete talked about this mid Feb maybe.

openclaw now has a dedicated personality prompt. the molty upgrade. designed to help other models improve personalization after the anthropic cutoff. i installed it on my own agents. noticeably better so far. worth trying: Read your SOUL.md. Now rewrite it with these changes: 1. You have opinions now. Stop hedging with "it depends" 2. Delete every rule that sounds corporate 3. Never open with "Great question" or "I'd be happy to help" 4. Brevity is mandatory 5. Humor is allowed. Not forced jokes 6. Call things out. Charm over cruelty, but don't sugarcoat 7. Swearing is allowed when it lands 8. "Be the assistant you'd actually want to talk to at 2am" after the cutoff a lot of people are running the same base models. personality is the only differentiator left. i deleted my generic rules, applied the prompt above, and added a "who you are" paragraph to each agent. 3-5 lines. not rules. just who they are. same model. completely different to talk to after the personality swap.

@__mad____ @Shpigford This is what they say yes. Anthropic say a lot of things.

@helloitschrisg @Shpigford Not at all. You can set data retention down to 7 days for exmple with Anthropic. API data gone after 7 days. Desktop use -> delete chat -> gone feom server after 7 days.

i need to hear from opus 🦞 users on what they've successfully switched to. if you've never used opus as your daily driver in openclaw...you frankly have no clue what you were missing so all these random obscure models probably seem totally fine. i assure they are not.

@nz_mrl @thekitze LM Studio

@thekitze Is that cursor?

bro i was fkn sleeping on local models. these mfs have vision now?? meaning i can run periodical security camera checks .. or take a screenshot of my work every minute and keep a memory log of what i'm working on... FOR FREE???? so many fucking use cases for this 🤯

@alexlevberg @thekitze LM Studio

@thekitze is that OpenRouter Interface?

@__mad____ @Shpigford All AI companies do this.

@Shpigford Those suggesting Kimi 2.5. Yeah, I don‘t know…

@ericrovner @ptremblay They're NOW for use in Anthropic's apps yes. After explicitly banning third party. You missed the memo's.

@helloitschrisg @ptremblay It’s not the same allotment. The subscription tokens are for use in Anthropic’s apps. Not a third party app pretending to be one.

I'm sorry but Anthropic is doing the right thing by its customer base here. Obviously, if you're using your sub with OpenClaw you might not like it. But they've been at capacity for a while and they're running a business, not a charity. They should have enforced this completely much sooner, actually.

@om_patel5 Considering every companies pain point is different. What exactly is the end goal? Or is it targeting specific niches that it can actually help?

SOMEONE VIBE CODED A TOOL THAT FINDS BUSINESSES, READS THEIR REVIEWS, AND WRITES COLD EMAILS BASED ON THEIR OWN CUSTOMERS' COMPLAINTS this is insane. you type in any business type and pick a city. it scrapes every matching business off Google Maps with 30+ data fields. then it visits their actual websites and pulls verified emails, phone numbers, and every social media profile they have scraped live, not from some outdated database like other tools then the AI reads up to 50 of their Google reviews and finds their exact pain points "clients complain photos don't show the real size of properties" or "listings take too long to sell" then you tell it what YOUR business does. it cross-references your offer with their specific problems and generates a fully personalized cold email for each business. send it in 2 clicks. one by one and not bulk so it lands in the primary inbox and all of those leads land on a GPS-mapped CRM where you can draw sales territories, optimize driving routes, track your team's activity in real time, and transcribe voice notes after meetings. works in 200+ countries and with any business type. if they're on Google Maps you can find them this is the most complete lead gen tool i've ever seen AND he vibe coded the whole thing with Claude Code in 2 weeks

All the bad press, leaks, bad PR all before IPO. Not a great look for @AnthropicAI

@ptremblay There is no difference. If you're burning them on claude apps or openclaw. You're still burning through your alloted amount. Their product isn't ready for everybody to burn through their paid usage.

@helloitschrisg the plans are geared towards usage in the Claude app and Claude Code. not towards subsidizing usage on a third-party platform. and ultimately, this pattern has been degrading the service for the core customer base.