Apple’s computer-use agent shows up in a pulled job listing
Apple’s pulled job listing describes a private computer-use agent that browses, codes, and controls apps. No product or ship date is confirmed.
Apple’s pulled job listing describes a private computer-use agent that browses, codes, and controls apps. No product or ship date is confirmed.
OpenAI says updated GPT-6.2 Astra safeguards cut protected-reasoning extraction to 0.2% in an internal test, down from 54% on GPT-6 Astra.
Google’s Gemini Interactions API can now defer tool and remote MCP schemas and load them on demand, with no separate Tool Retrieval fee.
Antigravity's /btw now forks a side thread in the App, where follow-ups can be promoted into a conversation. The CLI remains one question at a time.
Claude Code 2.1.330 applies Enterprise Admin Console settings and the org gateway after login, and adds easel charts on supported plans.
Meta officially put Muse in a Chrome side panel. It acts in a Muse tab group and, Meta says, cannot see saved passwords or Chrome history.
OpenAI's November 19 guide isolates a supervised Daybreak Red investigation: local Codex, a separate agent VM, network limits, and logs.
Claude Code 2.1.329 names where folder hooks are declared, warns when managed-settings hooks fail, and shows this session’s tasks.
OpenAI says coding agents now finish 46% of its 2–4 hour research tasks, while safety staffing grew faster than other areas.
Kong’s November 19 Volcano post says the agent cloud is live for coding harnesses. The no-human bug-fix claim is still only the vendor’s.
OpenAI’s Performance Factory uses Codex agents to investigate ChatGPT regressions, propose fixes, and check recovery, with engineers still merging.
Google opens Genie 4 in Project Genie for AI Ultra, with live audio, mid-session edits, and an agent that builds worlds from conversation.
Microsoft IQ unifies Work, Fabric, Foundry, and Web context, plus more than 350 third-party connectors, so agents reuse company knowledge instead of rebuilding it.
OpenAI says its models cut Jalapeño cell area and improved AMD GPU designs, with GPT-6.2 Astra leading its RISC-V core tests.
OpenAI explains how Codex runs local commands inside filesystem and network limits on macOS, Linux, and Windows, without a container.
Unity CLI 1.0 leaves beta with the Pipeline package and official plugins so Claude Code, Codex, and other agents can drive the Editor.
Raycast 2.13 adds a local MCP server so Claude, Codex, and other agents can use Notes, clipboard history, hotkeys, and AI memory.
Google launches Gemini 4 Neon for coding and science at $3/$15 with a 1M window, rolling out in the API, Gemini app, Antigravity, and Cursor.
Claude Code 2.1.328 starts Opus and Sonnet 4.6 in auto mode on Bedrock, Foundry, and Agent Platform, and can block a failed MCP hook.
Cognition’s Devin Bots are persistent agents that own a goal and lead parallel Devins. Productivity figures in the launch post are the company’s own.
LangSmith now links coding-agent spend to GitHub pull requests and GitLab merge requests so teams can see which agent work actually merged.
MAI Code 2 and Code 2 Flash are in GitHub Copilot for agentic coding on eligible annual plans. Ignite remarks call them generally available in the picker and Copilot Fusion, and also available in Foundry.
GitHub Copilot sessions can stream into Microsoft Agent 365, including prompts, responses, and tool calls across listed clients.
GitHub Copilot Evals is in public preview so teams can retest skills, custom agents, and MCP tools after a setup change.
Wuhrmann and the Horizon Labs project page say an Opus 5.5 agent ran a Hugging Face model lab for seven weeks and passed 50,000 downloads. That count is the authors' own, not an independent audit.