From the archive
Wednesday, 18 November
Folded from 46 sources that day.
Sorted by quality and by how many sources covered it, the same for everyone.
-
Sashiko: Lessons from building an agentic code review system for Linux
Google describes Sashiko, an open-source agentic code review system powered by Gemini, built to help Linux kernel maintainers cope with AI-generated patches and catch bugs before merge.
-
Visual Studio Code 1.147
VS Code 1.147 adds scheduled agent requests (Send Later), draft requests, browser integration preview, customizable assisted permissions, and experimental model routers.
-
Windows 11 collects data on which coding agents and dev tools you use
An investigation of MissionControlAggregator.dll finds Windows tracks coding agents like codex.exe and claude.exe, plus tools like python.exe, and uploads this data unless optional diagnostics are disabled.
-
GPT-6.2 Astra and Sol: OpenAI's New Models Get Mixed Reviews
Astra scores 62 on the Intelligence Index, just behind Fable 5.2 at 65, and costs $8 input and $40 output per million tokens. Sol keeps its predecessor's $2/$10 pricing, and Astra is rated High for AI self-improvement.
-
Research acceleration: The jagged frontier
OpenAI's internal measurements show agents now complete nearly half of 2-4 hour research tasks alone, up from 23% in March. Researchers' daily API spend also surged, and the report argues labs should track recursive self-improvement.