✍️ Blog
-
Your Grafana, Now Talks to Agents: The Azure Managed Grafana Remote MCP Server
Every Azure Managed Grafana instance now ships with a built-in remote MCP endpoint. Point Copilot, Claude, or a Foundry agent at it and they can search dashboards, query Prometheus, run KQL against Kusto, pull Application Insights traces, and run automated health checks — no plugins to install, no servers to...
Read more → -
Sorting Networks Part 2: From 27/28 to a Source Generator for Sizes 2–64
Taking a hardcoded SIMD sort for two sizes and turning it into a C# source generator that handles sizes 2–64, any type, with zero allocations — shipped as a NuGet package. 80 PRs total, all built with AI.
Read more → -
From Ten Days to Ten Minutes: Reviewing a Half-Million-Line Skia Bump
Bumping SkiaSharp to a new Skia release used to mean staring at a 500,000-line diff for the better part of two weeks. Now I open one HTML report, click around for ten minutes, and I'm done. Here's how a small AI skill changed my life.
Read more → -
Sorting Networks: From Paper to Performance with AI
Implementing a new CS paper algorithm in C# using GitHub Copilot, then benchmarking against .NET's built-in sort — achieving up to 41x faster performance with SIMD vectorization. 59 PRs in 3 days, from research paper to optimized library.
Read more → -
Multi-Model Adversarial Code Review with GitHub Agentic Workflows
What if your code reviewer never got tired, never missed a pattern, and always checked with two other experts before flagging an issue? We built an automated code review system that dispatches three AI models in parallel, runs adversarial consensus to filter noise, and posts findings directly on your PR....
Read more →