.NET

@dot.net

Free. Cross-platform. Open source. A developer platform for building all your apps. Run by the .NET Community team at Microsoft.

🚨 Build broke? VS Code’s new MSBuild Binlog Analyzer turns chaotic .binlog noise into answers — root causes, perf hotspots, diffs, regressions, even one‑click fixes 😎 👉 Blog: ift.tt/7HjkrhP

MSBuild Binlog Analyzer promo: large headline "Analyze MSBuild Binary Logs with Copilot in VS Code" on a purple background beside a VS Code screenshot showing a BinLog Explorer sidebar, build comparison metrics, and lists of errors and warnings.

SQL MCP Server = safe AI access to real data. 🤖🔐 LLMs give intent, SQL MCP builds deterministic T‑SQL. On‑behalf‑of auth = no “shovel with its own identity.” Seven tools, no hallucination soup. Embeddings + semantic search coming next. 👉 Watch: buff.ly/iurdpGB

Promotional banner for "The Modern .NET Show" with the title at top, a smiling host holding a glass of orange juice on the left, a purple robot mascot holding a microphone on the right, and blurred colored source code filling the background.

Agent skill testing, but make it controlled chaos! 😅 Try this: 🧪 Deterministic mocks 🔁 Stable responses 💸 No API spend 🛡️ Dev Proxy = real URLs, fake payloads 🚫 Zero prod side‑effects 👉 Read: buff.ly/hvmxTE6

Slide titled "How to test agent skills without hitting real APIs" explaining evaluating agent skills using mocked API calls, showing a cloud "/api" shield icon and a circular refresh icon on a purple–pink gradient.

#dotNETRocks chats with Jimmy Bogard about the decision to go commercial and what he's learned after a year. Making customer transactions is part of the process for sure, as is a new array of issues and ideas. Listen in... 👂️ buff.ly/iAQV5gY

Promotional podcast cover for Microsoft's .NET Rocks! episode "Commercial Open Source Update" showing two enthusiastic hosts—one leaning into a vintage microphone, both smiling and pointing at the viewer—purple panel with Microsoft logo and a microphone icon.

AX evals: no more ✨vibes‑based✨ scoring. Clear criteria 🎯, real skip rules 🚦, calibration loops 🔁, and environment fidelity that keeps results trustworthy. If you want AX signal instead of noise, this is the recipe. 👉 Read: developer.microsoft.com/blog/buildin...

Bild

Your agent eval has vibes you didn’t intend. 😅 OS shell swaps change behavior, file paths bias decisions (“azureuser” strikes again), and LSP feedback silently fixes code before you ever see errors. 👉 Read the blog: buff.ly/ottV1Oh

Bild

MCP Live lands Sept 9 🎉 A free, deep‑tech livestream on MCP servers, agent tooling, VS Code builds, GitHub MCP, FastMCP, DCR/EMA auth, and more. If you’re shipping agents or wiring MCP into enterprise tools… this one’s 🔥. 👉 Read + register: buff.ly/OeMMDZR

MCP Live! title in large blue‑purple type on a dark background, with a floating stylized UI panel and cursor hovering over stacked purple and blue 3D bars, suggesting a live tech demo or presentation

Modernizing .NET? Here’s your map. 🗺️ Upgrade when support ends, security hits, or perf walls appear. Follow the STS/LTS rhythm. Then containerize, cloud‑shift, and let the Copilot modernization agent guide the heavy lifting. 👉 buff.ly/eV34eXc

Promotional graphic encouraging readers to "Read the docs" for upgrading .NET apps, showing a stylized code window and code icons on a purple–pink gradient background.

Hot take 🔥: Don’t rewrite your CLI for --json. Args = perfect correctness across models JSON = shell‑escaping fails, retries, $$$ (4–11× cost) Smaller models especially need the constraints. 👉 Blog: buff.ly/Odt1cbF

Title slide reading "Don't Rewrite Your CLI for Agents" with a purple "Agent Experience" badge and subtitle "Why args beat JSON on correctness and cost", accompanied by an example terminal command icon and cracked curly braces symbolizing broken JSON.

🎨 .NET upgrades now come with a live canvas → Assess → Plan → Execute → Track code + build failures in real time Copilot’s upgrade agent makes modernization visual, structured, and way less chaotic 😎 👉 Read the blog: buff.ly/PbFwaA6

Promotional banner urging developers to modernize .NET within the GitHub Copilot app, with bold headline on a dark gradient background and two circular icons: a Copilot pilot-helmet logo and a purple .NET square.

Model upgrades aren’t automatic wins. Sonnet 5 used 12× more tokens on architecture tasks and scored lower on idiomatic quality (78% vs. 90%). But on SPFx upgrades, it hit 100% completion and followed instructions precisely. Measure first. Upgrade second. 👉 buff.ly/GQinE4W

Warning that upgrading AI models isn’t always better: large headline reads “Not all model upgrades are upgrades.” A side-by-side chart compares v1 (purple) with higher agent-experience score and lower cost, to v2 (red) showing higher cost but lower agent experience.

MCP C# SDK v2.0 landed 🎉 Stateless-by-default MRTR for interactive tools Smart HTTP headers ASP.NET Core-native A tidy, practical upgrade for MCP builders. 👉 Blog: ift.tt/Fx7gHVG

Announcing v2.0 of the official MCP C# SDK, shown as bold white text on a purple-to-pink gradient background with a large stylized exclamation mark on the left.

📊 AI benchmarks can fool you — models ace public tests but stumble in real agent workflows. Harnesses, extensions, SDKs… they all shift results. If you care about actual performance, this breakdown is gold. 📚 Read the blog: buff.ly/jrT6vqz

Slide titled "What AI Benchmarks Are Not Telling You" with subtitle "Why leaderboard scores don't predict your outcomes," showing a 92% trophy icon and a bar chart contrasting high benchmark bars with a lower real-world result.

How has testing changed in the era of AI? #dotNETRocks and Debbie O'Brien discuss using LLMs to help build and validate tests, using testing tools like Playwright to validate application workflows and generate documentation, and how docs are more important than ever. Listen in... 👂️ buff.ly/vkQSEg5

Promotional podcast cover for a .NET Rocks! Microsoft episode titled "The Role of AI in Testing," showing two enthusiastic hosts by a vintage microphone gesturing and pointing toward the audience.

AI agents don’t start blank — they start with a plan. If your docs don’t contradict that plan, they get ignored. The move: explicitly say the default approach fails, then show the correct one. Instant plan swap. Cleaner outputs. Happier pipelines. 👉 buff.ly/toszbHe

Slide titled "Your Agent Already Has a Plan" about agent experience, showing a friendly robot holding a checklist and a caution barricade icon on a purple-pink gradient, with subtext: "Why telling agents what to do isn't enough when they've already decided."

Models haven’t seen your .NET SDK — so they guess. 😅 This post shows how to spot the “closest‑match trap,” run baselines, and teach your .NET API from scratch with layered bootstrapping. Perfect for custom libraries, internal frameworks, and enterprise .NET stacks. 👉 Full scoop: buff.ly/Z4uuOr6

Slide title "When the model has never seen your code" explaining bootstrapping agent knowledge for proprietary code with zero pre-training data; visuals show a brain with a question mark and a locked code document, indicating unknown model knowledge and protected code.

A whole .NET modernization journey — now beginner‑friendly. Free, open‑source course that upgrades a real legacy ASP.NET app to .NET 10 using GitHub Copilot’s modernization agent. 👉 Explore the course: ift.tt/QwA0VIj #dotNET #ASPNET

Announcement banner reading "Announcing .NET Modernization for Beginners" on a purple-to-pink gradient background with a large stylized exclamation mark on the left.

In case you missed the news... .NET 11 Preview 6 dropped and it’s packed! ⚡ Runtime perf wins, NativeAOT boosts, async validation, OpenAPI 3.2, Blazor Virtualize upgrades, MAUI CollectionView2, EF Core + F# improvements, and more. 👉 Check it out: buff.ly/RLtgxi5

.NET 11 Preview 6 announcement: large text on a deep-blue underwater background with a purple cartoon submarine shining a headlight and trailing bubbles.

🧠🔥 .NET dev truth bomb: models don’t have preferences — they have context vibes. Ask inside your real .NET repo and the answer flips to your stack instantly. It’s not opinion… it’s pattern‑matching. Feed context, get clarity. 👉 Blog: buff.ly/cdzSWCg

Headline "Models Don't Have Preferences, They Have Context" with subtext explaining LLM preference evals measure your context, not the model; illustration of connected speech bubbles and documents representing prompts and context.

⚡🐳 Copilot + Testcontainers = .NET turbo mode. Copilot drafts your tests instantly, Testcontainers spins up real SQL deps in CI/CD for production‑level confidence. Write less, test smarter, ship faster. 👉 Read more @ThePracticalDev: buff.ly/pSmqbpF #dotNET #CICD

Promotional banner for a DEV.to article titled "Supercharging .NET Development with GitHub Copilot and Testcontainers," featuring a "Read the article" call-to-action over an abstract purple geometric background.

🧠🔥 .NET dev PSA: stop feeding your agent a whole API textbook. It already knows your CRUD + auth flows — bloated skills just burn context and tank accuracy. Measure the baseline, find the gaps, build the cheat sheet, not the manual. 👉 Blog: buff.ly/TxzJVhg #dotNET

Bold headline "Stop Overloading Your Skills" with subtext "The model already knows. Stop wasting tokens." showing a cluttered tablet overflowing papers transforming into a compact phone icon, illustrating condensing agent instructions to save tokens.

AI agents need audit trails. This .NET Core guide shows: 🔵 Append‑only Blob audit logs for every governance event 🟣 OpenTelemetry → App Insights for metrics + traces 🟡 Dual output on every policy check + violation 👉 Blog: buff.ly/rxBKd9T

Auditing and Telemetry for the Agent Governance Toolkit — a two-panel slide showing an AI agent and governance warnings on the left, and an architecture/data-flow diagram on the right illustrating audit logs written to Blob storage and telemetry sent to Application Insights for .NET Core.

🧠⚙️ Visual Studio’s built‑in Agent Skills are getting seriously clever. They now bring .NET + Azure expertise straight into the IDE, with project‑aware reasoning and smarter automation. 👉 Dive in: buff.ly/9QfBbgk

Promotional graphic showing Visual Studio's built-in Agent Skills for .NET and Azure, featuring a screenshot of the Agent Skills settings panel on the left and the headline "Built-in Agent Skills bring .NET and Azure expertise into Visual Studio" on the right.

〰️ HAPPENING TODAY at 10AM PT 〰️ Join the ASP.NET Core team for a tour of what's new for #ASPNET Core and Blazor in .NET 11 Preview 6: Async minimal API validation, automatic CSRF protection, and more. See you LIVE here ➡️ buff.ly/X3NgTUO #dotNET

Banner for ASP.NET Core Community Standup titled "What's new in .NET 11 Preview 6?" showing two circular headshots and a line listing the event hosts.

Let's talk AI Diagnostics. #VisualStudio is leveling up: AI‑powered diagnostics, Copilot agents that debug + profile, and a future where accessibility APIs let AI drive apps. Andy Sterland’s insights = 🔥 for devs. 👉 Watch: buff.ly/lgV1Fiq

Podcast episode title "Episode 408 — AI DIAGNOSTICS with guest" displayed above a split-screen video call showing two hosts wearing headphones and glasses speaking to camera.

July .NET servicing updates are out 🔐 🔧 Fixes across .NET 10/9/8 🛡️ CVEs addressed 📦 Updated installers + containers 📘 New release notes Blog: ift.tt/0YxwhW3

Announcement banner reading "July 2026" with headline ".NET & .NET Framework Updates," featuring a purple .NET logo in a white circle and scattered pink/purple geometric shapes.

CoreCLR now runs all .NET MAUI mobile apps. 🔥 iOS/Mac Catalyst = faster than Mono, Android ~10% within. Full .NET diagnostics on device + early NativeAOT groundwork. 🧪📱 👉 Blog: buff.ly/SigMcMc #dotNET

CoreCLR Progress Report for .NET MAUI headline announcing updates, with a "dotnet/maui" label and subtext about better tooling, app sizes, and performance on a purple gradient background.

In this #MergeConflict podcast, we dive deep into: ✔️ why users expect consistency ✔️ how small details create polish ✔️ and why understanding your users' mental models beats flashy design every time ...all with real-world examples. Listen in... 👂️ buff.ly/FRmqvxH

Podcast cover for Merge Conflict — episode "UI.md: Opinionated design rules for coding agents" with hosts credited on the left; dark left panel lists the topic and the right features a bold MERGE CONFLICT title and podcast logo.