Microsoft is testing a Health tab in Copilot with connectors for Fitbit, Garmin, Oura, Apple Health and Health Records.

Manus AI has officially announced its new always-on Agent with long-term memory, connectors and access to Manus capabilities.

Only telegram for now 👀

Google tests Hatter agent and App Store tools in Stitch update

Google’s Stitch design platform is testing a new “Hatter” agent aimed at advanced, multi-step UI creation, alongside features for automatic app store asset generation and native MCP export, signaling deeper AI-driven design workflows for developers and product teams.

🗞 #stitch

Google is testing a new Hatter agent on Stitch, along with a new tool to generate App Store assets and a simpler way to setup MCP connector for coding tools like Gemini CLI, Claude Code, Antigravity, Cursor and others.

"Create high-quality designs with Hatter agent"

Microsoft tests Researcher and Analyst agents in Copilot Tasks

Microsoft is developing a unified “Tasks” feature for Copilot, combining Researcher, Analyst, and a new Auto mode into one interface with scheduling options. It could enable complex, automated workflows across Microsoft 365, Windows, and Edge, though no release date is set.

🗞 #microsoftcopilot

BREAKING 🚨: Microsoft is working on new Researcher and Analyst agents for Copilot.

Both will be available in the upcoming Tasks feature, where users can schedule and execute complex agentic prompts.

OpenAI is working on a new model configuration setting that lets you set the default model and thinking effort.

Notion is rolling out an image generation feature for Notion AI. Which model do you think this is?

Did you get it too? 👀

Everything is a Computer 👀

"Perplexity Computer"

BREAKING 🚨: XAI IS ROLLING OUT GROK 4.20 (BETA), POWERED BY 4 AGENTS ACROSS ALL APPS!

Did you test it already? 👀

4 Grok agents are named as: Grok, Harper, Benjamin, and Lucas!

Communication between agents can be seen in the thought process.

BREAKING 🚨: ANTHROPIC STARTED ROLLING OUT CLAUDE SONNET 4.6 TO THE USERS! CLAUDE SONNET 4.6 SCORES 72.5% ON OSWORLD COMPUTER USE BENCHMARK.

It is expected to be a good model for Cowork tasks, with performance close to Opus 4.6 at a lower price.

BREAKING 🚨: CLAUDE SONNET 4.6 COMES WITH 1M CONTEXT WINDOW IN BETA!

"Claude Sonnet 4.6 is our most capable Sonnet model yet. It’s a full upgrade of the model’s skills across coding, computer use, long-context reasoning, agent planning, knowledge work, and design."

Google released the "Revise" feature for Slides on NotebookLM, allowing users to customise individual screens with a prompt.

Anthropic releases Claude Sonnet 4.6 with 1M context for all users

Anthropic has released Claude Sonnet 4.6, offering near‑Opus‑level performance with a 1M token context window in beta, stronger coding and reasoning, improved reliability, and broad availability across all Claude plans at unchanged pricing.

🗞 #claude

Cursor launches long-running agents for Ultra+ users

Cursor has released its long-running agents to Ultra, Teams, and Enterprise users, enabling autonomous execution of complex, multi-day software tasks. Powered by a custom harness and multi-model validation, the system delivers production-ready code with minimal oversight.

🗞 #cursor

Control browser on Comet is now powered by the recently released Claude Sonnet 4.6 model on Perplexity Pro accounts.