Tenobrus @tenobrus

Ranked #1 on Backscroll (2026-07-09) — surfaced for substance, not engagement.

swe 1.7 just launched and it's looking pretty great vs the paretos for frontier models here... and composer 2.5 is looking pretty terrible. on *cognition's* benchmark suite. meanwhile in composer's launch post, it looks pretty pareto-great on *cursor's* benchmark suite. imo this

How it ranks: Backscroll reads my Twitter/X timeline and ranks every tweet for substance with an LLM taste rubric — not raw engagement. Curated by Surya Dantuluri.