Simple Explanation of LLM's, Coding Agents

      GitHub Copilot, Claude Code, and Cursor are not AI models!  They are Coding Agents/Assistants.

They are the developer experience layer that sits in your IDE and seamlessly accesses various Large Language Models (LLMs) such as Claud Opus/Fable/Sonnet/Haiku or GPT5.

The actual intelligence comes from the underlying Large Language Model (LLM) such as GPT-5, Claude Opus/Sonnet, Gemini, 

Naming breakdown:

  VS Code / Visual Studio /Claude Code (IDE)    

      ─────▼────

GitHub Copilot / Cursor Extension / Claude Code Extension        (Coding Agent)      

      ─────▼────

 GPT-5 / Claude Opus / Gemini (LLMs)

Local vs Premium LLMS: 

Cursor can reference online LLMs like Claude Opus and use Ollama to run Qwen3-Coder (my pref. for now), DeepSeek-Coder, Llama 3.x, Mistral.  Ollama lets me point to a local LLM and use it for free (saving on tokens; there are other advantages as well).

My Setup for local LLM usage:

VS Code                                   +

Cursor Extension                     +

Ollama                                     +

Local LLM (Qwen3-Coder)    = 🚀 AI-Powered Software Development

Azure Container Jobs with Docker containing E2E Playwright testing

Overview: I recently did a great project with Playwright to continuously test Canvas Apps.  This post outlines how I did it.

Reporting: Every test suite run and the tests inside are documented in SharePoint lists. p Below you can see for a project called feedback the tests that verify DTAP Canvas apps (Dev, Test, and Prod).

The Feedback app, in production, is showing the availability tests run recently

 

CI: I decided to use Azure Container Jobs to run the Playwright tests on a Docker image.

Jobs: The Docker image get params and starts the type of tests, the trigger uses cron timing.

Azure Container Job: Each time a job is called, a new instance is created. This means multiple jobs can run simultaneously, and on each job instance I get multiple COUs, so I spawn out 2-4 Playwright processes so the tests run faster.

Client-Side vs Server-Side Rendering Websites

Web applications generally fall into the CSR or SSR.

Feature	Client-Side Rendering (CSR)	Server-Side Rendering (SSR)
Rendering location	Browser	Server
Initial page load	Slower	Faster
SEO	Worse	Better
Server workload	Lower	Higher
Browser workload	Higher	Lower
Typical use	Internal apps, dashboards	Public websites, e-commerce
System Classification	Power Apps Canvas App OutSystems (Low Code) Mendix (Low Code) React SPA Flutter Blazor WebAssembly Angular Vue	Power Pages Next.js Website ASP.NET MVC PHP E-commerce Product Page Blazor Server Ruby on Rails Razor Pages

The table is worth understanding as it affects architectural decisions as the project progresses.

Code Apps (Power Platform)

Code Apps (Power Platform) look to be gaining momentum. I think they will work for corporate app development as they offer governance and ALM as part of the Power Platform.   I saw this summary and thought it was a great overview.

AI infiltrates Development Segments so quickly

A decade ago, the companies that used Agile processes along these lines built the best software using CI/CD, and Automated Testing were the big winners. 

5 Years ago, low-code and platforms provided governance and faster ways to build apps.

We are seeing AI coming into all 6 steps, and low code is optional; it still offers good governance for businesses.

Test
Was: Unit Tests, Postman, Spec Flow, Selenium 
Becoming: Postman, MCP, Unit test generation

Deploy
Was: Scripts, PowerShell, TeamCity, Ansible, ARM, Portal UIs, TFS, Git, Terraform, Pulumi
Becoming: BICEP, MCP, VS Code/IDEs, GitHub, Claude CLI, GitHub Copilot

Code
Was: Java, C#, NodeJS, SDKs, RUST, GO APIs, Low-code/No-code
Becoming: Python, TypeScript, Power Platform, Agents

Key Anthropic terminology for Ai (and Claude)

Overview: I have found that many people don't understand what Anthropic and Claude do, how they relate to each other, or how they fit into the AI industry.

The key takeaway is that Claude has two main parts: Tooling (e.g., Claude AI chat, Claude code, and Claude Cowork), and models for AI. The main product lines for these LLMs are Opus, Sonnet, and Haiku.

VSCode with Claude Code or GitHup Copilot

Overview: I have seen people really struggling with understanding that Claude is not GitHub Copilot (GHCP).  And GitHub Copilot is not an LLM. 

Terminology: 

Claude by Anthropic is made up of various parts, and it helps to be more specific.  Calusde is amazing at providing great Large Language Models (LLMs).  There are Claude Opus (for programming the lastest us 4.7), Claud Sonnet, and Claude Haiku.  GPT-5.4 is OpenAI's current flagship LLM.  Gemini 3.1 Pro is known for UI-focused coding. I found Gemini 3.0 good, but I don't use it that often.   

Claude also offers other services beyond supplying LLMs like Claude Code and Claude CLI.  These are the key ones for me:

Claude in Chrome — a browsing agent

Cowork — a desktop tool for non-developers to automate file and task management, rival to Microsoft 365 Copilot.

GitHub Copilot for VS Code is the equivalent of Claude Code for VS Code

Developing in VS Code:

In the screen below, I am using both Claude Code and GitHub Copilot in the VS Code IDE.

Here I have some C# code that creates and deploys an Azure Function. I can use either option until I run out of credits with my monthly (GHCP) or hourly (Claude) subscription allowance.  When I go over my GHCP allowance, I have it set up to use my Azure Credits.