X

@cursor_ai

Cursor agents from your phone: you can now check in, reply to, or start tasks for agents running on your computer from the Cursor iOS app; the agent keeps running if the phone loses signal. post (https://x.com/cursor_ai/status/2107618653701296162) · 1,796❤ · 125 RT

@poteto

Release and QA automation: built two Grok Bot Team Bots, sandcastle as release manager and poteto as engineer bot, in a Slack channel that runs release cuts and QA. post (https://x.com/poteto/status/2107527180263829827) · 2,142❤ · 81 RT

Grok Bot backend change: quotes the announcement that Grok Bot will route each task to the best backend model, including Claude Opus 5.5, MidJourney, and Suno. post (https://x.com/poteto/status/2107728968501903652) · 1,876❤ · 56 RT

First and last mile: says her favorite Grok Bot use is the first mile (connecting apps to work out what to do) and the last mile of getting work done. post (https://x.com/poteto/status/2107510472601985336) · 1,256❤ · 58 RT

@rauchg

AI Gateway confidence fallback: quotes Vercel's beta feature that reruns an agent decision on a fallback model when the primary model's confidence drops below a threshold. post (https://x.com/rauchg/status/2107606246350266469) · 202❤ · 6 RT

GTA quip: jokes that at this rate AI will get us GTA 7 before GTA 6. post (https://x.com/rauchg/status/2107449561451274688) · 6,046❤ · 232 RT

@simonw

Mistral pelican: Mistral Large 4 passes his pelican-on-a-bicycle SVG test. post (https://x.com/simonw/status/2107479817780400577) · 706❤ · 26 RT

LLM music: asks when models became able to compose competent music after Opus 5.5 wrote game music better than he expected. post (https://x.com/simonw/status/2107496114299822393) · 270❤ · 15 RT

New LLM plugin: released a plugin for his LLM tool that sends prompts to OpenAI's new Jev-like decision model based on GPT-6 Luna. post (https://x.com/simonw/status/2107630159927062921) · 176❤ · 4 RT

@HamelHusain

Similarity metrics for evals: similar wording does not show whether an answer works; check specific failures instead, though similarity helps for retrieval and output diversity. post (https://x.com/HamelHusain/status/2107486073987219784) · 45❤ · 4 RT

Anti-slop job post: points out a job ad for an anti-slop writer that was itself written with AI slop. post (https://x.com/HamelHusain/status/2107699140134363331) · 22❤ · 0 RT

@swyx

Coding agent poll: asks what people's default coding agent is in October 2026. post (https://x.com/swyx/status/2107646238585950540) · 34❤ · 2 RT

@hwchase17

LangChain hiring: quotes Paraform's Talent Density Index ranking and invites people to join LangChain. post (https://x.com/hwchase17/status/2107766611646038216) · 12❤ · 1 RT

Silent

@martinfowler, @karpathy: no original posts in window

Gaps

@xAI: API not authorized to read this account's posts (skipped)

GitHub trending

tester-army/e2e (https://github.com/tester-army/e2e): Next generation e2e testing framework for web and mobile apps. (TypeScript, +1,725★ today)
mattpocock/skills (https://github.com/mattpocock/skills): Skills for Real Engineers. Straight from my .agents directory. (Shell, +889★ today)
earthtojake/text-to-cad (https://github.com/earthtojake/text-to-cad): Give your agent CAD superpowers. (Python, +619★ today)
boykopovar/AnyPS5 (https://github.com/boykopovar/AnyPS5): Tool for automatic PS5 executables porting to Linux and Windows. (C++, +949★ today)
pbakaus/impeccable (https://github.com/pbakaus/impeccable): The design language that makes your AI harness better at design. (JavaScript, +616★ today)