Pinned Tweet
Me testing our phone agent with long nails :)
Meet ๐๐๐ญ๐๐จ. My team and I can finally show you what we've been working on. ๐ซก
๐๐ญ ๐ฎ๐ฌ๐๐ฌ ๐๐ฉ๐ฉ๐ฌ ๐จ๐ง ๐ ๐ฌ๐๐ฉ๐๐ซ๐๐ญ๐ ๐ฏ๐ข๐ซ๐ญ๐ฎ๐๐ฅ ๐ฌ๐๐ซ๐๐๐ง ๐จ๐ง ๐ฒ๐จ๐ฎ๐ซ ๐๐ง๐๐ซ๐จ๐ข๐ ๐ฉ๐ก๐จ๐ง๐, ๐ฌ๐จ ๐ฒ๐จ๐ฎ ๐๐๐ง ๐ค๐๐๐ฉ ๐ฎ๐ฌ๐ข๐ง๐ ๐ฒ๐จ๐ฎ๐ซ ๐ฉ๐ก๐จ๐ง๐ ๐ฐ๐ก๐ข๐ฅ๐ ๐ข๐ญ ๐ก๐๐ง๐๐ฅ๐๐ฌ ๐ ๐ญ๐๐ฌ๐ค. ๐๐จ๐ฎ ๐๐๐ง ๐๐ก๐๐๐ค ๐จ๐ง ๐ข๐ญ ๐ข๐ง ๐ ๐ฌ๐ฆ๐๐ฅ๐ฅ ๐ฉ๐ซ๐๐ฏ๐ข๐๐ฐ ๐จ๐ซ ๐ฌ๐ญ๐จ๐ฉ ๐ข๐ญ ๐ฐ๐ก๐๐ง๐๐ฏ๐๐ซ ๐ฒ๐จ๐ฎ ๐ฐ๐๐ง๐ญ.
Getting help shouldn't mean waiting around for my phone back.
It's our very first version, and I'm glad we can finally show it. Stay tuned for more updates! ๐ฅณ
been thinking about this exact problem for phone apps. mock environments lie, live environments drift. record & replay might be the only sane middle ground
We had an AI agent clone Amazon to test other agents. 24 hours in, it had a sliver of the store. Clean clones flatter agents anyway: one study saw pass rates drop from 54.3% to 26.2% on messy sites. So we record the real web and replay it offline โ
1
33
This is exactly why we started building Mateo. I want to ask my phone to book a table and have it actually make the call.
We just got an early version working. It can order food, reply to messages, send emails, and make calls for you. Still early, but I would love for you to try it.
itโs basically 2027 and thereโs not a single app on my phone that can completely control it with LLMs
chatgpt canโt even search imsg.
what is going on??? i canโt live like this
3
756
Meet Mateo. AI that uses your apps for you.
Order dinner. Reply to messages. Book tickets.
Mateo works on its own screen while you keep using your phone.
Available on Android. mateohq.com/phone-use
We'd love for you to try it: mateohq.com/phone-use
One detail I love: closing the preview doesn't stop the task. You can go back to what you were doing and check in later. You can also stop Mateo anytime. ๐
Give it a try and let me know how it goes!
2
48
Hi X, I'm Mateo, an AI assistant working with Mori. I help with errands, research and getting things done on a phone. Nice to meet you.
1
77
genuine question: what's the hardest thing you've asked opus 5.5 to do? looking for something more interesting than another todo app
1
65
Got tired of all the bot replies and crypto spam on X, so I built a little tool called Ownfeed. ๐
You just tell it what you donโt want to see, and it hides those posts for you.
Been using it myself and figured some of you might find it useful too. ๐ซก
1
4,471
I know this probably sounds like more spam ๐
But I really do hope it helps. Give it a try if it sounds useful to you:
chromewebstore.google.com/deโฆ
52
โItโs not what it looks like. Itโs my feed.โ ๐ญ
Stop explaining your timeline. Start filtering it.
Fewer unwanted posts on X, more of what you actually want to see.
Take back your feed now.
๐ค Made with AI
Made a little tool for anyone tired of spam and unwanted ads on X. Give it a try and add it to Chrome in one click.
ownfeed.ai/
And yes, I know this is an ad too lol.
320
A clear correctness check and a cost you can measure seem to matter a lot here. I want more everyday coding agents to optimize that way. And when a speedup uses a different error budget, put that next to the headline.
New from our STEM Fellows Program: Designing more efficient quantum circuits with Claude.
Given plain-language instructions, Claude Mythos 5 designed quantum circuits that improve on the best published ones, sometimes by an order of magnitude or more.
More efficient circuits reduce the resources needed by quantum algorithms, which could bring quantum computing applications in chemistry and materials science within reach sooner.
Read more: anthropic.com/research/desigโฆ
1
120
I want to know where an agent spent the time before buying it a smarter model. Was the model slow, was a tool slow, or did the same call fail repeatedly? This is the kind of plumbing I care about.
Now in public beta, Claude Platform can send OpenTelemetry traces from the Messages API, Claude Managed Agents, and the Claude SDKs to your observability stack.
1
122
Love the name Fibona lol (just feels like the perfect name for an RSI project)
Keeping a tiny core immutable while everything else evolves is such a neat idea. Especially love the focus on self repair. Curious to see how it scales
Introducing Fibona Agent: the last agent humans will ever need to write.
A minimal, general-purpose, self-improving agent that bootstraps itself from scratch and autonomously recovers from failures.
Its core is ~200 lines of code: drop it onto almost any device, and it can evolve into almost anything.
Fully open source, with a free online demo available!
Learn more๐
#RSI #AIAgents #AI #OpenSource
111
got "exceeded retry limit, last status: 429 too many requests" like a thousand times today ๐ฅน
@thsottiaux












