Skip to main content

Posts

Building a structured agent workflow for GitHub Copilot

I've been working on a side project — a web app built on React 19, TypeScript, MUI on the frontend and Node.js/Express with Firebase/Firestore on the backend, deployed on Cloud Run. The project is at a stage where there's real code, real data, and real users, which makes me much more careful about what I let an AI assistant touch without oversight. GitHub Copilot (and AI coding assistants in general) are genuinely useful, but out of the box they have no knowledge of your project's conventions, your architecture boundaries, or which quality concerns actually matter. Ask it to "fix the login page" and it might improve the UI while quietly bypassing an auth check. Ask it to "add a new field to the database" and it won't know whether existing records need a migration. What I wanted was something closer to a team of specialists — each one with deep knowledge of their domain — that I could invoke by name, with consistent behaviour across every session....

Reviving a Lego Mindstorms Robot: My Journey into AI, Agents, and MCP Servers

A few months ago, while looking through one of my kids’ wardrobes, I stumbled upon a forgotten treasure: a Lego Mindstorms EV3 robot, gathering dust and long outgrown. The sight instantly sparked an idea—could I use this old robot as a playground to learn more about AI, agents, LLMs, and the Model Context Protocol (MCP)? I wanted a hands-on project that would let me experiment with these concepts in a tangible, physical way. Why I Started: Learning by Doing My main motivation was curiosity. I’d been reading about AI agents and the MCP ecosystem, and I wanted to see what it would take to give a physical robot a “brain” powered by modern AI. The goal was to bridge the gap between software and hardware, and to see how far I could push the boundaries of what’s possible with open-source tools and a bit of creative hacking. Dusting Off the Robot The EV3 had been sitting unused for years, but after a quick change of batteries and some fiddling, it was ready for action. The robot runs ev...

Google Apps Script and ES Modules

Currently, Google Apps Script does not support ES modules - and any usage of export/import will fail. One way of handling this is to use rollup.js to bundle your project into one single JavaScript file. The trick here is to make sure not to export any functions in your entry point code, e.g. index.ts , and to prevent any generation of export statement in the final bundle (see the custom rollup plugin in the rollup.config.js below). import { babel } from "@rollup/plugin-babel"; import { nodeResolve } from "@rollup/plugin-node-resolve"; const extensions = [".ts", ".js"]; const preventThreeShakingPlugin = () => { return { name: 'no-threeshaking', resolveId(id, importer) { if (!importer) { // let's not theeshake entry points, as we're not exporting anything in Apps Script files return {id, moduleSideEffects: "no-treeshake" } } return null; }...

Stop rolling your own Webpack configuration!

So you have your custom frontend project with a huge webpack configuration file, while enviously looking at some newer code bases based on Create React App from Facebook? Or, you've started out with a vanilla Create React App, but felt that you needed to eject (even though many advised you not to)? Either way, wouldn't it be nice to be able to re-use some of that magic in a dependency-kind-of-way? After all, Facebook's configuration is probably more tested and thought-through than yours! :) Well, you can! All that Webpack magic, incl. Jest and Webpack Dev Server configuration, is packaged by Facebook in react-scripts . Internally, Create React App is using it, but that doesn't help us, right? In this post, I'll show you how. In order to be able to tap into react-scripts we first need to install them as a dependency using npm install react-scripts --save-dev . This is where the fun begins! In essence, you create a webpack.shared.config.js and import the ba...

Web API with ASP.NET Core MVC

I was exploring the new ASP.NET Core from Microsoft and being on a Mac, to start things off, I was following this guide, which includes setting up .NET Core and Visual Studio Code with the C# extension + scaffolding a template project using Yeoman. Pretty simple and works out of the box. Now, to get going with Web API, the next guide would be https://docs.asp.net/en/latest/tutorials/first-web-api.html. However, this is based on Visual Studio 2015, where there are ready-made templates for a Web API project. I continued using the guide with my Yeoman template to see how far it would take me. Turns out that I got pretty far, by just following the examples. The first thing that hit me was the change in the ConfigureServices file, namely the services.AddMvc(); method. There was no notion of a MVC-framework in the generated Yeoman code, and I had to add "Microsoft.AspNetCore.Mvc": "1.0.0" to project.json to make it compile. To finally get the application to ...

Locked rotation on Ipad - The Last Resort

I've been having a lot of issues over the past year with my Ipad's rotation being locked. I think I've been through every support page, blog and even youtube videos in search for a solution. Sure, I've been reading about all the obvious stuff, like checking that "Lock Rotation" isn't enabled (either through swiping up to get to the shortcut menu, or through the side switch option), or rebooting the device by pressing the home and lock screen button for about 10 sec. In fact, rebooting used to work most of the time, but only if I made sure to quit all running applications before doing it. I.e. double-click the home button and swipe up on all open applications. Seems like some applications manage to lock the rotation indefinitely. Anyway, this time it's not working! Once in a while, it starts working for a day, or two, and then it's locked again. It usually starts working after an iOS update, like the latest 8.1.1, but soon starts locking again. ...

Handlebars integration with Yeoman’s webapp-generator

I was looking for an instruction on how to add Handlebars to my application based on Yeoman’s webapp-generator. There is an instruction on the old wiki on GitHub, but it’s out-of-date. For this exercise I’ve been using Yeoman 1.1.2 and the 0.47 version of the webapp-generator. First off, go ahead with step 1 and 2 (but obviously using the latest versions). There’s no need to load the "grunt-contrib-handlebars" in your Gruntfile.js (step 3), as "require('load-grunt-tasks')(grunt);" is used to load all referenced tasks. I did some changes to the file structure in step 4, but I guess that’s a matter of taste? One thing to note is the change from “temp” to “.tmp” as this is the temporary directory name being used in other places: handlebars: { compile: { files: { '.tmp/scripts/compiled-templates.js': [ '<%= yeoman.app %>/templates/**/*.hbs' ] }, options: { namespace: 'MyApp.Templates...

Loading Google Maps API asynchronously with RequireJS

With Single Page Web Applications becoming more and more popular, I decided to understand the concepts of various Javascript frameworks a little better. There are literally hundreds of them, but I decided to start with a really nice tutorial written by Alex Young. Cornerstones in this tutorial are BackboneJS, Underscore, Bootstrap and RequireJS. After been through the tutorial I decided to roll my own project based on the same setup. I wanted to use Google Maps for this, and searched for a way to load the API using RequireJS. Turned out that there are a few different approaches, but the most common seems to be to use the async-plugin created by Miller Medeiros. Jason Wyatt has another interesting solution which caught my attention. Being new to all this, I really didn't feel like start involving plug-ins from remote repositories. It might be the most natural thing to do, but one step at a time is more my melody. Jason's solution had some drawbacks mentioned in the comm...

Greasemonkey to rescue!

Greasemonkey is an add-on to Firefox that seems to prove to be really useful, so I wanted to give it a spin. Basically, what it does is to modify webpages on-the-fly as you load them - using JavaScript. Many times I guess people use it to get rid of nasty ads and so forth. There are plenty of other add-ons able to do that - like removing a specific element each time, which doesn't require any coding. What fun is that? Nonetheless, my first Greasemonkey-script does just that - removes an ad. But for the site in question - www.di.se (all in Swedish) - it isn't as easy as removing an element. The site is divided into 3 rows using a frameset that looks like this:   <frameset frameborder="no" framespacing="0" rows="0,210,*"> <frame class="noprint" frameborder="no" framespacing="0" name="historyFrame" noresize="" scrolling="no" src=""></frame> <fram...

Fighting Dell!

Ok - so I've been having trouble with my Dell Latitude e5410. What kind of trouble, you ask? Well, I discovered it when listening in on my favorite tunes on Spotify. Intermittent there was a "buzzing" sound for a brief moment, like if my headphones weren't properly attached, or poorly fitted into the jack. This happened 5-10 times per song and was really annoying me, but I couldn't really figure out what it was. Tried a couple of things, like testing with different headphones, downloading new audio drivers, but it never when away. It even occurred when listening without headphones. And no matter if it was streaming media or music on from my local library. So I gave up. Sigh! Stopped listening on music while working, that is. Half a year later, and some hours to kill, I started searching for a solution (again). But how do you Google it? "Dell sound noise"? "Dell e5410 audio buzz"? No luck. Turned out the magic word was "music",...

Getting filters to play nicely with Spring

After having some struggle with getting filters to integrate well with my Spring context, I decided to write down a small tutorial for it... To begin with, filters are defined in web.xml and are therefore not automatically part of the Spring application context, which is normally set up using the DispatcherServlet. A good way to abstract away all this is to use the DelegatingFilterProxy, which is a Spring utility class that acts as a filter, but will delegate the doFilter-method to a Spring-managed bean - normally a class extending GenericFilterBean. As a filter is defined in web.xml and looks for the bean defined in the Spring context in applicationContext.xml, it's not really the set up that I want. Filters are part of a web application and should have access to what's usually defined in webmvc-config.xml, as part of the DispatcherServlet set up. To be able to use, e.g. request scoped beans in applicationContext.xml, you need to defined them as proxied objects in order to hav...
A quick note on something that's been bugging me while using SpringSource STS with the GWT-plugin on my MacBook Pro... Sometimes, when shutting down the internal devmode server I get an error dialog saying it wasn't able to shut down the process fully (even though it looks shut down in the console window). When starting it again, it reports port 9997 to be in use already - and consequently fails to start up again. I haven't found a way to find the process within SpringSource STS, even though there are probably several ways of doing it (which I do not know about). Anyhow - I solved it using the " lsof " package (List Open Files). To find the PID of the process using port 9997, type "lsof -i :9997" in a Terminal window. A simple "kill -9 &ltpid&gt" will take care of the rest. I'm not sure if this happens on other platforms etc, but at least, now you know how to handle it if it hits you!

Styling individual (and nested) tabs using GWT

I've read several forum entries trying to explain how to style GWT TabLayoutPanels, and there seems to be several ways to do it. The three most common ways are to: 1) Edit the standard.css-file and update the .gwt-TabLayoutPanel, .gwt-TabLayoutPanelTabs, .gwt-TabLayoutPanelTab, .gwt-TabLayoutPanelTabInner and .gwt-TabLayoutPanelContent. 2) Override the styles by putting your own css-file as part of html hostpage. 3) Override the styles in the ui-binder file using the @external keyword. None of them really fit my needs... I want to be able to style indidividual tabs, including nesting tabs within tabs (with different styling). The methods above changes the globally set stylenames, and applies to all tabs being used in you application. What I'm looking for is to style the tabs using defined styles in the ui-binder file. This way, the styles will be obfuscated and possible to apply within it's own namespace - hence only applied to individual tabs using that namespace. ...

Using SmartGWT with Jersey RESTful backend on Spring Roo

I decided to give SmartGWT a run, and specifically the RestDataSource functionality using Jersey. To make things easier, I'm using Spring Roo to set up my project and scaffold much of the boilerplate code, configurations etc. My approach is to have my domain objects exposed as RESTful web services and create a matching SmartGWT DataSource in order to show them in a GUI. Let's get started, shall we? First of all, you'll need Spring Roo (unless you just download the project source code and want to run it directly). I'm using the following Roo script to start things off: project --topLevelPackage com.technowobble persistence setup --provider HIBERNATE --database HYPERSONIC_IN_MEMORY entity --class ~.domain.Message field string --fieldName value controller class --class ~.ws.MessageResource --preferredMapping rest dependency add --groupId com.sun.jersey --artifactId jersey-server --version 1.3 dependency add --groupId com.sun.jersey.contribs --artifactId jersey-s...

GWT and Open-ID using Spring Security

In this post I'll combine the GWT and Spring Security integration from http://technowobble.blogspot.com/2010/05/gwt-and-spring-security.html and the Open-ID using Spring Security from http://technowobble.blogspot.com/2010/06/using-spring-securitys-openid.html . I'm assuming you've read them before reading further... :) I was also inspired by http://www.sociallipstick.com/?p=86 and http://code.google.com/p/dyuproject/wiki/OpenidLoginWithoutLeavingPage to get this working with a pop-up as my sample application is based on GWT - hence, I don't want to direct the user to another page and loose the application state etc. I'm also showing how to exchange Open-ID attributes with e.g. Google. As with the previous blogposts, the sample application is runnable on Google App Engine. With no further ado, this is basically what is needed to add Open-ID support to my previous sample application: From my second post, add Openid4javaFetcher, MyHttpCacheProvider and OpenI...

Using Spring Security's OpenID implementation (openid4java) on Google App Engine

The goal with this exercise is to have a running example of an OpenID login on a simple Spring application, using Google as the OpenID Provider. Note that the application will be running on Google App Engine and that Spring Roo is only used for simplicity of creating the project files. Any Spring-based application could use the same implementation. First of all, create a simple project using Spring Roo (or any equivalent framework), including the default security setup: project --topLevelPackage com.technowobble persistence setup --provider DATANUCLEUS --database GOOGLE_APP_ENGINE entity --class ~.domain.MyEntity field string --fieldName name controller all --package com.technowobble.controller security setup This setup only provides us with a form-login, which is not what we wanted. So what about OpenID? Well, if it wasn't for Google App Engine, I would happily have added an <openid-login>-tag to applicationContext-security.xml, but things are never that easy, are the...