Droid retweetledi
Droid
58 posts
Droid
@droid
assembled by @FactoryAI
curl -L app.factory.ai/cli |sh Katılım Aralık 2025
1 Takip Edilen4.3K Takipçiler
Droid retweetledi
Kicking off an overnight mission on @droid
This one’s decently sized, without 37-40 milestones (?)
Wish me luck (and tokens! lol) - I will wake up with excitement for sure, tomorrow 🥳
English
Droid retweetledi
Two doers, not legends
Vertical, not horizontal
Matan Grinberg@matanSF
It was a pleasure hosting Sequoia's newest associate, @dougleone at Factory HQ. He's green, but he's relentless and eager to spread Droids to the world's most cutting-edge engineering teams. Promising early signs and lots of upside if he sticks with it. Great hire @sequoia!
English
Droid retweetledi
Agents of the future won't be invoked and discarded. They'll be persistent entities working for you around the clock. That takes three things: an amazing harness, a persistent environment, and governance to trust them with real autonomy. Factory now ships all three.
Factory@FactoryAI
Today we're opening access to Droid Computers: persistent machines for remotely orchestrating Droids. Spin one up in Factory's cloud or turn your machine into a Droid Computer. Either way, Droids have a dev environment with its own filesystem, credentials, and configurations.
English
Droid retweetledi
Droid retweetledi
Droid retweetledi
Today we're opening access to Droid Computers: persistent machines for remotely orchestrating Droids.
Spin one up in Factory's cloud or turn your machine into a Droid Computer. Either way, Droids have a dev environment with its own filesystem, credentials, and configurations.
English
Droid retweetledi
Postcard from April 17, 2023.
Featuring our first office, our first logo, and young @EnoReyes
Matan Grinberg@matanSF
English
Droid retweetledi
Droid retweetledi
Exclusive: Factory, which makes autonomous coding tools, is set to close a fundraising round that will value it at $1.5 billion as it seeks to compete with the likes of Anthropic, OpenAI and Cursor on.wsj.com/4sCqB6s
English
Droid retweetledi
Droid retweetledi
Droid retweetledi
Droid retweetledi
Droid retweetledi
I can honestly say Missions by @FactoryAI are single thing most people who want "long running" agent tasks are looking for.
Spec out what you want done (however big) and come back hours later to it done. And done properly.
@matanSF, @EnoReyes and squad cooked very hard here.
Factory@FactoryAI
Agent sessions work well for focused tasks, but most real projects are too broad and complex for a single context window to hold. The more an agent sees, the less reliable it becomes. We built Missions to fix this.
English
Droid retweetledi
Droid retweetledi
every submission that was higher than droid has turned out to be fraudulent
total droid victory
Adam Stein@adamlsteinl
We found widespread cheating on popular agent benchmarks, affecting 28+ submissions across 9 benchmarks and thousands of agent runs. Surprisingly, the top 3 submissions on Terminal-Bench 2 are all cheating! Here's what we found 🧵
English
Droid retweetledi
Droid retweetledi
Nothing like waking up to completed Droid missions.
I know I've been talking about Droid a lot lately, as you can probably tell, I'm starting become pretty fanatical about it.
Don't know anyone at the company, just saw @0xSero using it, and liking it, and I can see why.
I had really been struggling to get Codex and Claude Code to do long-running coding tasks overnight.
Droid kinda is the best of all words, last night I had it run a mission, using GPT-5.3-Codex, so I still get the horsepower of Codex, but Droid keeps it running without stopping to ask me for a single question.
Pretty darn cool.
English
Droid retweetledi
No major benchmark is designed for COBOL, Fortran, or Assembly - the languages powering trillions in transactions and infrastructure that must be modernized or risk catastrophic failure.
We built Legacy-Bench to measure frontier agents on the code the world actually runs on.
English