Robert Cuadra

I write code. Opinions are my own.

[Robert Cuadra @robertcuadra]

Love this so much!

[Josh Pigford @Shpigford]

Knockoff is now live!

Filter out the knockoff crap brands on Amazon.

Sorry to brands like WNPETHOME, EHEYCIGA, YXYL, LU&MN, JOYIN, TOMY, GODONLIF, YOOJEE, LINGTENG, LANEIGE, VISCOO, BIODANCE, COOFANDY, BALENNZ, TOSY and LUENX.

https://t.co/9mLk0EAsfG

[Barstool Sports @barstoolsports]

Mexico City’s reaction to their second goal is something right out of a documentary

[Justine Moore @venturetwins]

Me using Claude Opus 4.8 to rename a file

[dex @dexhorthy]

someone hit me up about the new "claude dynamic workflows" feature, claiming "see, multi-agent works"

But really, the launch of this feature proves the exact point that I made back in June of 2025, along with @walden_yan, @tobi, @karpathy, and many others:

Deterministic workflows orchestrating small agent loops beats non-deterministic multi-agent or "agent soup" systems every dang time

everything is context engineering

[Karthik @karthikponna19]

"it worked on production just like it did on localhost"

[Cormac @cormachayden_]

software engineers before vs after agents

[Robert Cuadra @robertcuadra]

The only thing scary about that is people still haven’t figured out you shouldn’t give agents full access to things with API keys. It is not hard to spin up a separate agent to build a proxy for any service you want with logical guardrails to prevent disasters like this.

[Ryan Carson @ryancarson]

Scary.

Read this and make sure you harden your db backups!

[Robert Cuadra @robertcuadra]

@lifeofjer Step 1: don’t give an agent that kind of direct access. It’s not hard to setup your own guardrails via proxies for your agents.

[Robert Cuadra @robertcuadra]

@BillAckman @X AGREE!!!

[Robert Cuadra @robertcuadra]

🤣🤣🤣 I know this pain

[Jeff Escalante @jescalan]

Transition to gpt in openclaw not going so great this far 😂

[Ryan Carson @ryancarson]

The cost to run a truly useful Chief of Staff @openclaw on Opus 4.6 is $100-200 per day on the API.

[Robert Cuadra @robertcuadra]

Time will tell if this was a smart move for Anthropic. Maybe they think they're going to release something better? Maybe they know they're going to force everyone to OpenAI and cause them to burn a TON of money subsidizing.

What is shitty is doing it late afternoon on a Friday with basically 18 hour notice. Fortunately for me it's a one line change in ansible to update all of them.

I had several Claude MAX subs, and I just cancelled all but 1. I know I'm not the only one.

[Peter Steinberger 🦞 @steipete]

woke up and my mentions are full of these

Both me and @davemorin tried to talk sense into Anthropic, best we managed was delaying this for a week.

Funny how timings match up, first they copy some popular features into their closed harness, then they lock out open source.

[Robert Cuadra @robertcuadra]

The companies and people who treat AI as a feature update are going to waste cycles optimizing a model that's already cracking. The ones who treat it as a fundamental shift, who are willing to rebuild how the work actually gets done, are the ones who'll still be standing when the dust settles.

This isn't a technology decision, it's a structural one, and structural decisions don't wait for you to feel ready.

[Robert Cuadra @robertcuadra]

Agentic AI is breaking the operating assumptions entire industries were built on. The assumptions about how the work gets done and what it should cost. These shifts used to happen every decade or so, now they are happening quarterly with the rapid increases in LLM capabilities.

[Robert Cuadra @robertcuadra]

AI is doing the same thing to knowledge work, and it's moving faster than any transition I've seen in a quarter century of building software. The shift isn't about the technology. It's about the assumption that work has to be organized, priced, and delivered the way it's always been. Per human. Per hour. Per seat. They make less sense every month as quality outcomes start to matter more than the path to get there.

[Robert Cuadra @robertcuadra]

Your mental model may be the bottleneck, the story you're telling yourself about what AI is for.

Update the story and everything else will follow from that.

[Robert Cuadra @robertcuadra]

The most dangerous place to be right now is productive with AI, because productive feels like you're keeping up. Things are faster and it feels like progress.

[Robert Cuadra @robertcuadra]

The people pulling ahead right now aren't the ones with the best prompts or the most tools. They're the ones who sat down, looked at how they actually work, and rebuilt it around what AI makes possible instead of what they were already doing.