Introducing EvoSkill v2: a system that automatically improves AI agent skills without…

This is a dev post classified by AGK Intelligence as AI dev tools (a launch), kept by the Dev Radar because it carries real work, not commentary.

Introducing EvoSkill v2: a system that automatically improves AI agent skills without changing the underlying model or runtime. It tests the agent, learns from its attempts, and rewrites its instructions to improve performance. Our new paper compares EvoSkil v2 with SkillOpt by @MSFTResearch and SkillBoost by @AlibabaGroup on held-out tasks, reporting: +23.6pts on τ3-bench banking with no skill, using Claude Haiku 5.5 +7.7pts over SkillOpt on SpreadsheetBench v2 +5.4pts over SkillOpt on τ3-bench banking

Posted by Sentient (503.6k followers) 3 h ago · 26 likes · 5.4k views · view the original post on X. Kept by the Dev Radar as AI dev tools.

More dev work like this

• As part of the most insane AIE lineup to end 2026, we're also excited to announce our… — @swyx
• The sprites crack me up a bit, but I love the outputs I'm getting using Deepseek EXL… — @loktar00
• Running several coding agents shouldn’t mean juggling several disconnected terminals. — @DanKornas
• Supervisor-Skills 这个项目已经攒到 7.9K+ Star，它干的事挺妙——把博导十年攒下的科研、顶会发表和审稿经验，直接"炼"成了一整套 AI Skills。 — @bkdgiffug
• YouTube firing on all cylinders. — @rileybrown
• Example here of invoking ce-compound after a session didn’t result in any forced docs to… — @trevin
• October was a big month for Claude Code 🍁 — @ClaudeDevs
• Does your coding agent try to do manual JS Date math for lunisolar holidays? 🌕 — @ChromiumDev

Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.3k posts from 5k X accounts over the last 21 days, 9.2k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-03 18:17 UTC. Full method.