@dabit3@DevinAI building a small tool to track ai coding tools - claude code, codex, cursor, etc.
happy to try devin cli and add it to the list as well
https://t.co/ofwXbKFgRV
Today we’re releasing DeepSWE, a new standard for agentic coding benchmarks.
On public leaderboards, top models often look relatively close in capability. DeepSWE shows where they actually diverge, reflecting the realistic experience of developers in their day-to-day work.
our team spent the day in the profiler and cooked something we're proud of - paste any tweet/link and in ~3 seconds the app:
→ reads the full content
→ detects the right asset out of 400+ on Hyperliquid
→ writes a clean thesis + reasoning
used to take 15s. now it flies. 5x faster, same quality.
and the best part is good narratives earn - every trade on yours pays you back 💸
come trade with us at @swaydotfun 🫡
Agents don't need types. They're perfectly capable of pulling off incredible refactorings without. Give them a linter and a test suite, and you have all you need. Token efficiency is where it's at.
we've been cooking this all week on @swaydotfun!
now you can see how the people you follow are actually playing the chart - who went long, who fumbled, who's quietly stacking
opinions are cheap, positions aren't.
flip on social on any asset and have a look 👀