Claude Code 2.1.281 adds MCP URL-mode elicitation on 2026-07-28 connections so servers can open a browser-based flow without leaving a stuck waiting dialog. claude plugin validate now flags silently dropped .mcp.json entries, undeclared ${user_config.*} refs, and insecure URLs.
GitHub’s Sept 23 changelog opens a dedicated Copilot code review settings page on every plan (including Business/Enterprise): auto review on PR create/coauthor/ready for review, optional auto review for drafts and new pushes, and a Lite/Balanced default effort. Enterprise admins can set one default effort for org owned repos (orgs/repos can still override)…
Cognition put OpenAI’s GPT 6 Sol and Luna live in Devin Desktop and CLI. On FrontierCode 1.1, Sol matches GPT 5.6 Sol at 61% lower cost per task and sits within four points of Astra at under half the cost; Luna is the cheapest model on the board at under $0.10 per task…
Claude Opus 5.5 is rolling out in GitHub Copilot for Pro+, Max, Business, and Enterprise across VS Code, JetBrains, Copilot CLI, the coding agent, and more. Early tests matched Opus 5 on task resolution with fewer steps and tokens; Enterprise/Business admins can gate it via model policy…
Xiaomi released MiMo V2.6 Pro and Flash as open weights (MIT) plus a Distill Qwen 9B research model, claiming Pro leads open models on Artificial Analysis at 46 while still trailing Claude Fable 5.1 and GPT 6 Astra. The drop also includes 7,000+ RL task environments, an end to end RL framework, and mini harnesses for multi harness agent training…
Anthropic says nearly 950 Claude agents spent 21 hours and 210 million tokens scanning DNA data before flagging a previously uncharacterized bacteriophage enzyme system for human review. The company compares the find to CRISPR machinery but says practical uses are still unclear; it’s the first result from its new wet lab…
Spotted something we missed? Start a thread.
