We've found it surprisingly useful for collaborating with colleagues on scoped fixes and features. It's now the top contributor on our repo.
CODE
-
Roguelike Game Showcases LLM Rules Engine as Art Project
By
–
I’ll be at this event and assuming August productivity stars align will be showing off a roguelike game I’m developing as something of an art project and tech exploration of LLMs as an evolving-during-play rules engine.
-
LLM GUI Demo: Dynamic Ephemeral Interfaces Over Static Design
By
–
Cool demo of a GUI for LLMs! Obviously it has a bit silly feel of a “horseless carriage” in that it exactly replicates conventional UI in the new paradigm, but the high level idea is to generate a completely ephemeral UI on demand depending on the specific task at hand. https://t.co/Xgh1iwDmJl
— Andrej Karpathy (@karpathy) 19 juin 2025Cool demo of a GUI for LLMs! Obviously it has a bit silly feel of a “horseless carriage” in that it exactly replicates conventional UI in the new paradigm, but the high level idea is to generate a completely ephemeral UI on demand depending on the specific task at hand.
-
5 Hackathons Hosted by Devpost with Prizes and Deadlines
By
–
Explore these 5 Hackathons hosted by @Devpost Agent Development Kit Hackathon with Google Cloud PRIZES: $50,000 DEADLINE: June 23 Build multi-agent AI systems with Agent Development Kit and Google Cloud
JOIN HACKATHON: https://
bit.ly/googlecloudmul
tiagenti
… AWS Breaking Barriers -
Join SF Hackathon: Kontext Development with BFL Team
By
–
Come join us in SF on Saturday to hack on Kontext with the BFL team
-
Claude AI can run continuously for 24 hours straight
By
–
You can have Claude code running for 24 hours straight too, if you want.
-
Spawning Subagents in Claude Code: A Hidden Feature
By
–
Btw, since people don’t seem to know this, you can literally spawn subagents in Claude Code just by asking. pic.twitter.com/dOz29G6qlL
— Pietro Schirano (@skirano) 19 juin 2025Btw, since people don’t seem to know this, you can literally spawn subagents in Claude Code just by asking.
-
GPT-4.1 Mini: Affordable AI Model for Budget Developers
By
–
GPT-4.1 Mini: The Intern (Budget-Friendly Coding) Cheaper, faster version of 4.1 Handles big tasks well, slight accuracy trade-off Great for developers on a budget handling extensive data or codebases.
-
GPT-4.1 Advanced Coding Model with Massive Context Window
By
–
GPT-4.1: The Coder (Context Master) Massive context (up to 1 million tokens via API) Precise coding and detailed instructions Excellent for coding tasks, handling large documents/transcripts, and detailed instruction-following.