A compute device, document cards, and translucent speech bubbles connected in front of a laptop.

Edge0 Part 2: Microsoft Learn and Apple Translation Setup

As of September 13, 2026, the sample in this series uses Edge0-35B, Microsoft Learn CLI, and the macOS Translation framework. Following the responsibilities established in Part 1, the inference server and Python agent host run in separate terminals. This article covers installing Edge0 and its model, starting the streaming server, searching official documentation, translating between Korean and English, and normalizing questions. The complete files are README.md, apple_translate.swift, and microsoft_expert_agent.py in the sample Gist. ...

September 14, 2026 · 6 min ·  rkttu
Glass blocks above a silver laptop and stacked storage devices representing local model inference.

Edge0 Part 1: A 35B Local LLM and Grounded Agent on a 16GB Mac

As of September 13, 2026, Edge0 provides an MLX backend for Apple Silicon and the Edge0-35B-A3B-preview model. I ran the model on a 16GB M2 MacBook Air, then built a technical expert agent that answers using Microsoft Learn references. This series covers the requirements for running a 35B-class model, changes in answers after adding references, Korean input and output, streaming measurements, and conditions for an internal PoC. The complete runnable code is available in the Edge0 Microsoft Technology Expert Agent Gist. ...

September 13, 2026 · 6 min ·  rkttu
C# on a light brown musical staff with branching nodes representing automated releases across platforms

Building an Eight-Platform Release Pipeline for C# with netcoredbg

On September 6, 2026, I completed the first release of C# (with netcoredbg) 2.148.23001 in the vscode-csharp-autobuild repository I maintain. I published packages for Windows, Linux, Alpine Linux, and macOS on both x64 and ARM64 to Open VSX, then confirmed that the public downloads matched the hashes of the validated VSIX files. The first release includes the source versions of the C# extension and netcoredbg, platform-specific packages, and validation records. This article covers the existing community’s contributions, the motivation for using AI in this project, compatibility validation across eight platforms, and recovery after interrupted publishing. The scheduled workflow now handles everything from source builds to verification of public downloads when it finds a new version. If a problem occurs, it stops publishing that candidate. I also describe a possible follow-up development tool informed by this release. ...

September 6, 2026 · 13 min ·  rkttu
Illustration of a configuration file passing through a state manager into user-scoped SDK and runtime components

dotnetup Preview: Managing SDKs and Runtimes from global.json

Microsoft introduced dotnetup, a new tool for managing user-level .NET installations, in the prerecorded Microsoft Build 2026 session OD804, “Simplifying .NET installs with .NET Up”. As of August 22, 2026, the release/dnup branch of the dotnet/sdk repository provides preview installation instructions and command reference documentation for Windows, macOS, and Linux. This article examines the behavior available in the current preview through the problem statement and design goals presented in the session. ...

August 23, 2026 · 12 min ·  rkttu
Bright purple illustration of dotnet bot flying toward a translucent number 11

Evaluating .NET 11 Adoption Before Release

Microsoft plans to release .NET 11 on November 10, 2026. As of August 23, 2026, .NET 11 Preview 7, released on August 11, is the latest build. Microsoft continues to refine the runtime, SDK, C# 15, ASP.NET Core, .NET MAUI, Entity Framework Core, and the rest of the product family. Features and behavior may still change before the final release. Microsoft’s .NET 11 Preview 7 announcement This article examines .NET 11 across five areas: the support cycle, hardware baseline, runtime and developer tooling, C# 15, and application frameworks. The focus is on changes that affect migration rather than the number of new features. ...

August 23, 2026 · 9 min ·  rkttu
Conceptual image of wiring OpenAI Codex CLI to any LLM with 50 lines of C#

Running OpenAI Codex CLI on Top of Claude, Gemini, or Llama — in 50 Lines of C#

OpenAI’s Codex CLI ships a great editor-agent UX — shell tools, apply_patch, plan tracking, all of it. The catch: as of February 2026, it only speaks the OpenAI Responses API. Chat Completion support was removed (the WireApi enum in codex-rs/model-provider-info/src/lib.rs now has only Responses), which leaves Chat-Completion-only endpoints — Ollama, LM Studio, your favourite Llama runner — locked out. This post walks through how I used .NET 10 file-based programs and the IChatClient abstraction from Microsoft.Extensions.AI to stand up a Responses-compatible server in a single 50-line C# file, with OpenRouter as the backend, so that Codex CLI can run on top of whichever model I feel like that day. ...

May 27, 2026 · 9 min ·  rkttu
Conceptual image of C# and Python meeting for machine learning interop

Calling Hugging Face Models from C#: Running Whisper, sentence-transformers, and Stable Diffusion with DotNetPy 0.6.0

Over the weekend I shipped 0.6.0 of DotNetPy, a small C# library that calls CPython’s C API directly to run Python inside a .NET app. This post walks through the three machine learning samples bundled with 0.6.0 — semantic search with sentence-transformers, speech recognition with Whisper, and text-to-image with Stable Diffusion Turbo — and explains how the same release was also validated on PEP 703 free-threaded CPython. Starting Point: You Only Have C#, but the Model Lives on Hugging Face Every few months the same pattern repeats. I need Whisper for subtitles, or a sentence-transformer for search, or occasionally something like Stable Diffusion — but the only tool in hand is C#. The usual workarounds all come with a decisive downside. ...

May 11, 2026 · 10 min ·  rkttu
Cloud platform architecture concept image

From RoleEntryPoint to FoundryCBAgent — A .NET Developer's View of Microsoft Foundry Hosted Agent Architecture

This article analyzes the design of Microsoft Foundry Hosted Agent, currently in public preview as of March 2026, within the lineage of .NET architecture. This is not a recommendation for production use at this point, but focuses on the design direction that enterprise AI architects should pay attention to. What It Means to Put an Agent in Production When deploying AI agents to production in an enterprise, there are currently two options. ...

March 17, 2026 · 14 min ·  rkttu
Concept image of code running from a single C# file

A New Genre in .NET: The Era of NuGet-Free Single File C# Coding

C# isn’t becoming as lightweight as a scripting language — scripting languages are becoming envious of how fast it’s getting. Introduction dotnet run file.cs — the so-called file-based app — introduced in .NET 10, lets you run C# code with just a single .cs file, without a .csproj. However, its current execution speed is roughly 1.5 seconds on Windows and 0.8 seconds on WSL2 for the first run. Compared to Python’s python script.py at around 50ms, it’s hard to call this “scripting” with a straight face. ...

March 16, 2026 · 12 min ·  rkttu
Conceptual image of inspecting code with a magnifying glass

How AI Agents Figure Out APIs When Building an IDE with a Brand-New .NET UI Framework

2026-02-08 — LibraStudio Dev Log #1 Background Why Another .NET IDE? The IDE options in the .NET ecosystem are actually quite limited. Visual Studio is powerful but Windows-only, and even the Community Edition has restrictions on commercial use. The VS Code + C# Dev Kit combo is great, but the fact that C# Dev Kit is proprietary doesn’t change. Ultimately, vendor lock-in occurs at critical points, and the entire tool chain and workflow built on top becomes dependent on a specific vendor’s decisions. ...

February 8, 2026 · 8 min ·  rkttu