AI News

Cursor introduces Rollouts for deployment verification

Rollouts write a monitoring plan and watch changes as they deploy, verifying deployments so regressions are caught before users see them.

GitHub Copilot app's rebuilt pull request view stays fast on extreme diffs

GitHub says the view remains fast and smooth for a pull request with 2,200 files, more than 1 million changed lines, and more than 400 inline comments.

Tencent Hy introduces WebCraftBench for AI-built web apps

WebCraftBench has agents use a live app, then scores aesthetics, usability, and whether the original request was met. Tencent Hy reports an 85.3% match with human preference on 197 validated pairs.

Cline says Stealth Bunny Alpha is now free

Cline says the stealth model is free in Cline, with multimodal input and a 1M-token context window. It also describes the model as having fast inference and strong coding capabilities.

NVIDIA's Skill2Env compiles public Agent Skills into 7,971 terminal tasks

A Codex planner and creator turn each skill into tested terminal workflows, using a rubric taken from that skill's own quality criteria. Outcome-only RL moved Qwen3.8-27B from 49.4% to 54.1% on Terminal-Bench 2.1, while adding the rubric reached 50.1% but won more trajectory comparisons.

Arena shows side-by-side outputs from Claude Opus 5.5 and GPT-6 Sol

Arena.ai says it has generated side-by-side outputs from Anthropic’s Claude Opus 5.5 and OpenAI’s GPT-6 Sol, and that scores for Claude Opus 5.5 are coming soon.

Lakebase Branching lets coding agents test database changes outside production

The short-lived environments capture both schema and data, so tests do not affect production. Workflows use the Databricks CLI and AGENTS.md to compare changes with the parent branch before approved updates move through CI/CD.

Claude Marketplace now lets users discover tools, agents and expert partners

Users can add connectors and plugins such as Slack and Notion, buy agents and products from companies such as Cursor and CrowdStrike, and scale with service partners such as Accenture and Deloitte.

ClaudeDevs reports making claude.ai 3x faster in two weeks

ClaudeDevs says it made claude.ai three times faster over two weeks and describes using Claude to measure, debug, and improve performance, with prompts and methods included.

YouTube users will soon be able to design custom feeds with conversational prompts

Neal Mohan said at MadeOnYouTube that YouTube announced more than 30 new features and products for creators, artists, and viewers. Users will soon be able to design dedicated homepage feeds for any mood or routine with simple conversational prompts.

MentalHealthBench covers everyday support through acute crises

OpenAI says MentalHealthBench is designed for the full range of mental health conversations people bring to AI, from everyday support to more acute crisis scenarios, rather than only emergencies.

OpenAI releases MentalHealthBench for realistic mental health conversations

OpenAI has openly released MentalHealthBench, a new benchmark built with input from more than 80 mental health clinicians. OpenAI says it shows frontier models continuing to improve in realistic mental health conversations.

Google Vids adds Gemini Omni 1.1 Flash, free for Google accounts

News from Google says Gemini Omni 1.1 Flash and a new suite of creative controls are now in Google Vids. Anyone with a Google account can create videos there for free, without professional editing skills.

Cline says its desktop app can be tried for free with a Kimi K3 promo

Cline links the offer to its Kimi K3 promo at cline.bot/desktop, without stating an end date or eligibility limits.

Cline Desktop v0.0.34 adds worktrees, GitHub PR status, and parallel subagents

Parallel tasks each run on a separate branch under ~/.cline/worktrees/<id>. The release also shows a branch’s pull request link, merge status, and CI checks when gh is available, and can assign several tasks to subagents in one session.

Runway is now available in DaVinci Resolve

Inside DaVinci Resolve, Runway can generate, edit, and upscale directly on the timeline without leaving the project.

Artificial Analysis reports 11 new points on its Intelligence Index cost frontier

MiMo-V2.6-Pro, Claude Opus 5.5, GPT-6 Luna, and GPT-6 Sol account for 11 new points on Artificial Analysis’s Intelligence Index versus cost-per-task Pareto frontier, across different reasoning efforts. Claude Opus 5.5 (max with fallback) scores 58 at $5.98 per task, the highest result on the index.