Thivy

Setting up Azure Chat is now as simple as running these Azure CLI commands: 1. Initialise: azd init -t microsoft/azurechat 2. Deploy: azd up Hopefully @savostrowski is proud of AZD UP as we are! github.com/microsoft/azur… #ai #azureopenai #openai #azureai #azurechat

Three ways to author .NET agent skills — file, class, or inline — all running through one provider. Add script execution + human approval and you’ve got a flexible, governed skill model for real apps. Blog: buff.ly/4jUlnBi

@jeffhollan Sooo good!! Thanks for the video walk through Jeff!

Excited to announce the new preview for Microsoft Foundry Agents 🎉! You can now build, run, and deploy your agent using any model, any framework, any harness in the cloud 🧑‍💻 - check out the demo below This is not just any cloud compute environment; it's an agent-optimized platform with: 🖥️ Persistent microVMs - securely scale up and down without losing context 🛠️ Built-in tools (1000+) 👀 Observability and evaluations 👷 Guardrails 🔐 Private networking... and more

Great to meet with @BecaGroup in Auckland and see how they are using Azure, Foundry, and their BEYON platform to make the New Zealand Geotechnical Database more accessible and useful. It is a powerful example of AI helping engineers access critical data faster and make better decisions, as they build more resilient infrastructure across New Zealand.

Been brewing this with @tommsmoro for a little while. Now you can install agent skills with GitHub CLI github.blog/changelog/2026…

gh skill adds commands to GitHub CLI to discover, install, manage, and publish AI agent skills from GitHub repos. • Pin skills to tags or commits to control updates and ensure supply chain security github.blog/changelog/2026…

The @GitHub Research folks released a "Rubber Duck" agent for the Copilot CLI. Automatically get a review from a model from a different AI family. And their data shows that yes, this does in fact help. Quite a bit.

Check out our announcement blog post vite.dev/blog/announcin…

So we're still building, but we are starting shipping a set of .NET skills from our team. Ones that we've used, helped internal teams, etc. hopefully to help your coding agents do more and you do less of boring (or harder) stuff. devblogs.microsoft.com/dotnet/extend-…

Folks often ask for a list of best practices partnering with agents in @code, and we've compiled a list here!

@pierceboggan @code Auto pilot all the way!!

New in @code Insiders: Autopilot to work autonomously on a task until complete, and simpler controls for managing approvals.

Agents, for real work. The latest @code release gives you better agent orchestration, extensibility, and continuity. Here's what's new: 🪝 Hooks support 🎯 Message steering and queueing 🌐 Agentic integrated browser 🧠 Shared memory And more...

Come and join me on.design 👀 on.design/claim?ref=thivy

Imagine waking up to calm... ✅ Issues triaged ✅ CI failures investigated + fixes ✅ 2 new PRs improving your tests Chores done, problems solved 🪄 Join us in shaping the future of repository automation with GitHub Agentic Workflows. github.blog/ai-and-ml/auto…

@pierceboggan wow wow!! This Epic! nice one team!

You told us you wanted features available in Insiders to VS Code stable, faster. We're moving towards weekly stable releases to bring top features to VS Code. New this week: Message queuing, steering, hooks, and skills as slash commands. Send us your feedback on these weekly releases to let us know what you like and how we can improve!

The GitHub Copilot CLI has been a blast to use for build out apps with agent swarms. I use it nearly every day standalone and also directly inside of Visual Studio Code. The one and only @cinnamon_msft just published a full LinkedIn Learning course! linkedin.com/learning/termi…

Will do a bigger thing on this later, but I've added 25 Claude Code skills for .NET developers to the Claude Code marketplace. Covers stuff I care about that are non-obvious (i.e. how do I test transactional emails in ASP .NET Core?) and drives at improving LLM output (types!)

@timneutkens What tool did you use? Copilot? Cursor? Claude code? Did the implementation of skill discovery vary between the tools?

100% of Next.js evals pass with a small 8KB change to the agents md and providing the Next.js documentation. Teach your agent about all Next.js features: npx @next/codemod@canary agents-md This will apply the change to your preferred config and download the relevant documentation for the installed version. The codemod approach will help us prove out the changes before bringing this into Next.js core as a feature. We've been researching the effectiveness of skills, agents md, and other approaches. While skills do solve more evals, it doesn't solve all evals. Whereas the agents md covers all evals. More details in this post: vercel.com/blog/agents-md… Great work and research by @gao_jude.

📢 Meet Playwright CLI — a SKILL-friendly way of the browser automation. Learn more at github.com/microsoft/play…. Happy testing!

The Copilot SDK is now in technical preview! • Programmatic Copilot CLI access with Node.js, Python, Go, and .NET SDKs • Start multi-turn conversations, build custom tools, control sessions, and more • Unified API across languages Try the SDKs ↓ github.blog/changelog/2026…