标签：#ev

I finally shipped the first version of my freelance landing page. The funny part? I spent weeks thinking I had a design problem. I didn't. I had a content problem. The layout is good enough. The copy is good enough. The CSS is good enough. What the page really needs now is more real projects. Instead of redesigning it again, I'm going to spend my time replacing concept work with actual client work as it comes in. Sometimes the next version isn't another refactor. It's simply experience. 🔗 https://lksvn.com.br/freelance/

Instead of immediately opening pull requests when newer dependency versions are released, Dependabot now waits three days before suggesting upgrades, thus increasing the likelihood that malicious releases are identified and removed before they can be integrated. By Sergio De Simone

Foreman is an agentic coder that runs as Kubernetes resources. You describe work as a Workload, it decomposes into tasks, agents running on your nodes pick them up, and a branch comes out the other end with something deterministic standing between that branch and your main. This is the walkthrough. Four objects to understand, an install, an agent, a verifier, and a real run. Every command and every output below is from a working cluster. The four objects Foreman is deliberately small. Almost everything you do is one of these. Agent is a worker definition: which model it talks to, which tools it may call, and what budget it gets. An Agent has a role , and the two that matter here are coder and verifier . Workload is the unit of work you actually author. It carries an intent, a repository, and which agents to use. AgenticTask is what a Workload decomposes into. You rarely write one by hand; you read them to see what is happening. FleetNode is a node that has advertised itself as able to run tasks. The scheduler matches a task's required capabilities against these. The shape of a run is: you apply a Workload, the controller synthesizes AgenticTasks, the scheduler routes each to a FleetNode whose agent can serve it, the agent runs the model in a loop with tools, and the result lands as a branch plus a verdict. The idea underneath it Worth stating plainly, because it shapes every design decision: the model is not trusted, and specifically its claim to have succeeded is not trusted. A coder agent finishes by calling a tool that says "I am done, verdict GO." Foreman treats that as a request, not a result. If the model says GO and produced no diff, the run is recorded as NO-GO. If the verifier's checks do not pass, the work does not land, no matter how confident the summary was. That is the difference between an agent that writes code and a system you can leave running. Everything else in this post is plumbing around that idea. Install Foreman ships as a Helm chart that dep

Parking and matching puzzles look almost insultingly simple from the outside. A few cars, a cramped lot, tap or drag to move them out. But if you've actually tried to build one that feels tight — no janky collision resolution, no ambiguous "why didn't that move register" moments, no lag once the board gets crowded — you know there's a real systems design problem hiding underneath what looks like a weekend project. I recently went through the architecture of a parking/matching hybrid template built in Unity and wanted to break down the core systems the way I'd want them explained if I were reskinning or extending one myself. This isn't a marketing post — it's a walkthrough of the actual mechanics: how vehicle movement and collision resolution work, how the match/clear pipeline is structured, how level data is separated from movement logic so hundreds of levels don't require touching code, and how monetization hooks slot into natural break points without polluting gameplay scripts. If you want to see the finished product this breakdown is loosely based on, there's a working template here: Park Match Unity Game Template . Everything below applies whether you're building something similar from scratch or extending an existing base. Why Parking Puzzles Are Harder Than They Look The core loop — tap or drag a vehicle, it exits along a valid path, the lot clears one piece at a time — is trivial to describe and genuinely fiddly to implement well. Four problems show up almost immediately once you move past a static mockup: Valid-move detection : how do you know, at any given moment, which vehicles can actually move given their orientation and the current board state? Path resolution : once a vehicle starts moving, how does it navigate around other vehicles and obstacles without clipping through them or getting stuck mid-animation? Match/clear logic : when does a vehicle actually "clear" the board — on reaching an exit, on matching color/type with another vehicle, or both — an

So I had this idea that kept nagging at me. Every AI chat app works the same way. You type something, the model returns text or markdown, the UI renders it as a nice formatted paragraph. That is fine if you want an answer. It is genuinely boring if you want to actually build something. What if the AI could respond with a working game board you could click? What if saying "make it Barbie themed" actually transformed the whole interface while you watched? What if "add a starfield in the background" dropped an animated canvas behind your chat in real time? I spent a few weeks building exactly that. I call it FlowChat . Here is the live version: https://flowchat-public.varshithvh.workers.dev And yes, someone immediately asked it to play Tic Tac Toe and then asked it to switch to an Oppenheimer theme mid-game. I could not be prouder. The Idea Normal AI chat: model returns markdown, client renders it as text. Simple, predictable, boring. FlowChat: model returns raw HTML with CSS and JavaScript, client injects it directly into the DOM using a streaming protocol built on the browser's native template system. That one change makes the entire experience different. You are not reading about a game. You are playing one. You are not reading about a Barbie color palette. You are sitting inside one. The AI does not just answer questions. It rebuilds the UI from its responses . What You Can Actually Do With It I want to give you a feel for what this means in practice before getting into the technical bits, because the demos are more interesting than any architecture diagram. Games : Ask it to build Tic Tac Toe. You get a playable board, click-to-move, an AI opponent, win detection. Ask for Connect 4. Ask for Snake. The game renders in the chat as an agent bubble with a form inside it. Each move submits to the LLM which processes it and updates only the cells that changed. Themes : Say "change to a Barbie theme". The model injects CSS overrides and the whole interface turns pink. Me

TL;DR AI editors add a login check to your API routes but skip the ownership check, so any logged-in user can read another user's data by changing the ID in the URL (CWE-639, IDOR). It happens because tutorials treat "authenticated" as if it means "authorized," and the AI learned from those tutorials. The fix is one line: scope every lookup to the current user instead of trusting a raw ID from the request. I asked Cursor to build an endpoint that returns an invoice by ID. It gave me clean code. Auth middleware on the route, a database lookup, a JSON response. It ran on the first try. Then I logged in as a different test user and changed the number at the end of the URL. Invoice #1001 belonged to someone else. I got the whole thing back: amount, line items, billing address. No error, no warning. Just another user's private data on my screen. That is IDOR, an Insecure Direct Object Reference, and it is one of the most common holes I find in AI-generated APIs. The frustrating part is that the code looks secure. It even has an auth check. It just checks the wrong thing. The Vulnerable Code The endpoint below is broken because it confirms you are logged in but never confirms the invoice is yours. findById takes the ID straight from the URL and returns whatever it finds. // CWE-639: authenticated, but no ownership check app . get ( ' /api/invoices/:id ' , authenticate , async ( req , res ) => { const invoice = await Invoice . findById ( req . params . id ); res . json ( invoice ); }); The authenticate middleware does its job. It proves the request comes from a real, logged-in user. What it does not prove is that this particular user has any right to invoice :id . Change the ID, get someone else's record. Increment it in a loop and you can walk the entire table. Why This Keeps Happening AI editors confuse authentication with authorization because almost every tutorial they trained on does the same thing. Authentication is "who are you." Authorization is "are you allowed to

This tutorial builds a starter "Hello World" style agent using Kotlin and the native Kotlin version of the Agent Development Kit (ADK). The full sample project is available on GitHub: xbill9 / adk-hello-world-kotlin Kotlin ADK and MCP Hello World This project is a runnable Kotlin Agent Development Kit (ADK) demo. A Kotlin LlmAgent uses Gemini to decide when to call a greet tool discovered from a local Kotlin Model Context Protocol (MCP) server. The project has two Gradle modules: agent : the Kotlin ADK agent, Gemini model configuration, MCP toolset, and interactive ReplRunner ; server : the Ktor MCP server that exposes greet . Technology Stack Kotlin: 2.3.0 Kotlin ADK SDK: com.google.adk:google-adk-kotlin-core (v0.6.0) MCP Kotlin SDK: io.modelcontextprotocol:kotlin-sdk-jvm (v0.8.1) Ktor Framework: 3.0.0 (Netty, SSE, ContentNegotiation, CORS) JDK: Java 25 Build System: Gradle 9.2.1 (Kotlin DSL) Prerequisites Java 25 A Gemini Developer API key The Gradle wrapper is included. Configure Gemini Create the local environment file: cp .env.example .env Set GOOGLE_API_KEY in .env , then load it: source ./set_env.sh The file is ignored by Git. Run the Demo Start the Kotlin MCP server in one… View on GitHub What Is Kotlin? Kotlin is a modern, statically typed programming language created by JetBrains. It runs on the Java Virtual Machine (JVM), works alongside existing Java libraries, and is widely used for Android, backend, and multiplatform development. Static typing is especially useful when building agents. Agent configuration, tool schemas, and tool results can all be checked by the compiler before a prompt reaches the model. Installing Java This sample uses Java 25 . If Java is not installed, SDKMAN! is a convenient way to install and switch between JDK versions on Linux and macOS: Home | SDKMAN! the Software Development Kit Manager SDKMAN! is a tool for managing parallel versions of multiple Software Development Kits on most Unix based systems. sdkman.io After installing

Github URL: https://github.com/SID-Technologies/Pilum In December 2025, I open-sourced Pilum , a multi-cloud deployment CLI that deploys to Cloud Run, Lambda, Azure, Cloudflare Pages, npm, Homebrew, and Docker Hub from a single pilum.yaml . The announcement post covered the architecture: recipes, ingredients, handlers, wave-based execution. That was the "it compiles and the tests pass" version. Three months and 40+ pull requests later, Pilum deploys all of SID Technologies , platform-core, Torch, Statio, every website, every npm package, and itself. This post is about everything that broke between "it works" and "it ships production software." ( Update: * there's now a seven-months-in addendum at the bottom. It's mostly about what stopped breaking.)* The Timeline Dec 2, 2025 : First commit. Baseline CLI with recipe system. (#1) Dec 4 : Homebrew release workflow. Pilum dogfoods its own deployment. (#2-#4) Dec 31 : Service graph, --only-changed , file embedding support. The "I need this for real" features. (#22-#24) Jan 3-12 : Documentation and bug fixes. The quiet "oh, this doesn't actually work" phase. (#25-#29) Feb 6-7 : The big feature sprint — 8 features in 48 hours. Wave deployments, npm recipe, Cloudflare Pages, Azure Container Apps, Cloud Run Jobs, environment variables, JSON output, history command. (#30-#38) Feb 8-12 : The big fix sprint — YAML parsing, package manager issues, build failures, error swallowing, GCP secrets, Cloudflare execution. Everything from the feature sprint broke something. (#42-#49) Feb 24 : Security hardening and npm publishing fixes. (#51-#53) Mar-Apr : Memory-based worker allocation, wave ordering bug, orchestrator rewrite. The "I thought this was done" phase. (#58-#62) The pattern is clear: features ship fast, fixes ship faster, and the real bugs show up a month later. Wave-Based Deployments Were Broken Wave-based deployment was the headline feature in #31. Services declare dependencies, Pilum builds a dependency graph, topological

Rust web development just got more interesting. If you prefer a video version: A few days ago, I spoke with Carl Lerche and Julien Scholz about Topcoat , a new batteries-included framework for building full-stack reactive web applications with Rust. Now I have finally tried it myself. Topcoat comes from the ecosystem behind Tokio and Axum, but it aims to provide a very different experience: routing, server-side rendering, reactive components, UI tooling, asset bundling, and hot reload in one framework. A simple Topcoat application This is what a basic application looks like: use topcoat ::{ Result , router ::{ Router , RouterBuilderDiscoverExt , page }, view ::{ component , view }, }; #[tokio::main] async fn main () { topcoat :: start ( Router :: builder () .discover () .build ()) .await .unwrap (); } #[page( "/" )] async fn home () -> Result { view! { <! DOCTYPE html > < html > < body > hello ( name : "World" ) </ body > </ html > } } #[component] async fn hello ( name : & str ) -> Result { view! { < h1 > "Hello, " ( name ) "!" </ h1 > } } The syntax feels surprisingly familiar if you have used server-rendered frameworks before. Pages and components are written in Rust, while the view! macro keeps the HTML structure easy to understand. What surprised me My first experience was better than expected. The setup was simple, the basic example was actually basic, and the development server provided working hot reload. That last part may sound normal to JavaScript developers, but it makes a huge difference for the Rust developer experience. Topcoat includes or plans to support: Server-side rendering Reactive components Module-based routing Tailwind integration Reusable UI components Asset bundling Fonts and icons Cookies and sessions Database integrations The experience feels closer to frameworks such as Laravel, Django, Rails, or Next.js, while allowing developers to build the application in Rust. Is Topcoat ready for production? Topcoat is still at an early stage. Break

Veo is launching accessibility-focused electric tricycles in Denver, with plans to put its e-trikes on the streets nationwide.

We run a React app and an Angular admin panel at work. Same company, same brand, and on paper the same design. On screen it was a different story. The React button had a 6px radius; the Angular one had 4px. The focus rings were two slightly different blues. Nobody noticed until somebody did. And every time design changed a token, someone got to hand-port it into two codebases. Twice the work, and it still drifted. So I went looking for something that treated both frameworks as equals. The React kits don't speak Angular. The Angular ones don't share a look with anything on the React side. Nothing let me define the design once and have it show up, the same, in both. So I built bpdm/ui . One rule: the look lives in tokens I gave myself one hard rule: nothing about how a component looks is allowed to live inside the React or Angular code. Colour, spacing, radius, the easing on transitions, all of it sits in @bpdm/tokens as plain CSS variables, and both framework packages just read from there. The component owns structure, behaviour, and the accessibility plumbing. The look comes from the tokens. @import "tailwindcss" ; @import "@bpdm/tokens/tokens.css" ; Change one token and both frameworks move together. There's no "now go sync the Angular theme" step, because there's only one theme to sync. Four ship in the box (two light, two dark). Override the variables and you've re-skinned all of it. The same component, twice React: import { Button , Badge } from " @bpdm/ui " ; export function Example () { return ( < Button variant = "primary" > Get started < Badge appearance = "soft" > New </ Badge > </ Button > ); } Angular: import { Component } from " @angular/core " ; import { BpdmButton } from " @bpdm/ng " ; @ Component ({ selector : " app-root " , imports : [ BpdmButton ], template : `<button bpdmButton>Get started</button>` , }) export class App {} Same padding, same radius, same focus ring. The accessibility isn't literally shared code: Radix does that work on the React s

Despite what influencers may say, you don’t need to spend $99.99 on Dyson’s HushJet Mini Cool or $149.99 for the Shark ChillPill to survive the summer whenever you leave the comfort of air-conditioning. My family has found all the comfort it needs to survive humid baseball games, sweltering concerts, and sweaty hikes with a couple […]

Cómo montar un motor de contenidos que no te arruine (julio 2026) Si sigues pagando 300 euros al mes por herramientas "todo en uno" de marketing, estás tirando el dinero. A mediados de 2026, la tecnología para automatizar ha bajado tanto de precio que los costes de infraestructura de contenidos son casi ridículos. La clave no es la herramienta cara, es conectar piezas pequeñas con APIs baratas. Aquí tienes cómo tengo montado mi flujo de trabajo ahora mismo. La pila tecnológica (el stack) Para automatizar sin gastar, olvida las plataformas de marketing tipo HubSpot o plataformas cerradas. Mi setup actual es este: Cerebro: Claude 3.5 Sonnet (vía API). Es mejor razonando que GPT-4o para tono editorial. Orquestador: n8n (corriendo en una VPS de 5 euros al mes en Hetzner). Base de datos: Notion (vía API para gestionar el calendario). Distribución: Ghost para el blog y la API de LinkedIn/X para el alcance. Coste total: Menos de 15 euros al mes. Paso 1: El disparador (el calendario en Notion) No uses un Excel. Usa una base de datos de Notion con cuatro columnas: Estado , Título , Prompt_Contexto y Fecha_Publicación . Cuando cambias el estado de "Borrador" a "Listo para generar", el webhook de n8n se dispara. Aquí es donde empieza el ahorro. No envías toda la base de datos, envías solo el registro nuevo. Paso 2: El prompt como código, no como texto La mayoría de la gente comete el error de pedirle a la IA: "escribe un post sobre X". Sale basura genérica. En 2026, si no das contexto, el contenido no posiciona ni recibe interacción. En tu nodo de n8n, construye el prompt de forma dinámica. Así es como envío la estructura a la API: { "model" : "claude-3-5-sonnet-20260620" , "system" : "Eres un redactor técnico senior especializado en SaaS B2B. Tu estilo es directo, sin paja, sin adjetivos innecesarios. Evitas los clichés de marketing de 2024. Tu objetivo es educar, no vender." , "messages" : [ { "role" : "user" , "content" : "Escribe un artículo corto basado en este punto clav

As a site reliability engineer in a global company, I'm running a modern (well, relatively modern, to be honest and modest) cloud-native stack: HashiCorp Vault as the secret manager, workloads on Kubernetes clusters in AWS (EKS), and development workflows automated through Jenkins (legacy) and GitLab CI. This setup is, quite likely, familiar to you — it's the normal playbook in the cloud-native era. In theory, we have the right tools for both security and efficiency: After all, we have a state-of-the-art secret manager integrated with everything. But in reality, it's far from the truth. See if you resonate with the following scenarios: Scenario A: A new colleague just joined the team. Manager: "Your initial password to log in to your corporate account came to me via email, but since you can't log in to your mail account just yet, here, take a picture of my screen." (In some companies, taking a picture of a computer monitor would get you fired, I'm not kidding.) Scenario B: A developer needs a temp password to access a database. Dev: "Where is the newly created temporary password? Need it for debugging." Ops: "In the Vault." Dev: "I can't access Vault." Ops: "No, you can't. It's not safe to open UI access to Vault. Corporate policy." Dev: "Then how can I get the password?" Ops: "Well... Technically, the password isn't in the Vault. There is a Jenkins pipeline that calls the Vault API to generate a temp password, then stores it in Jenkins secrets. You need to request access to the corresponding Jenkins pipeline, trigger it, then get the secrets from Jenkins." Dev: "Why on earth do we store secrets in Jenkins when we have Vault, which we aren't allowed to use?" Ops: "Corporate policy, just told you." Scenario C: A new ops team member needs to update a certificate for a service running in production for the first time. Ops: "Where is the old cert?" Mentor: "In K8s as a secret." Ops: "Where is the cluster?" Mentor: "In AWS." Ops: "How do I access that?" Mentor: "You need

Grafana Labs has expanded the capabilities of Grafana Assistant, enabling its AI-powered observability assistant to query and correlate data across more than 30 different data sources through natural language. By Craig Risi

A few months ago, my AI coding workflow looked something like...

Beeeep. Beeeep. Beeeeeeeeeeeeeeeep.

As it promised last year, Baidu has started testing its robotaxis in London along with partners Lyft and Freenow.

We're Byte Me , a software agency from Alkmaar, the Netherlands. The most valuable advice we give clients is usually not "Let's build something new"; it's "Let's not touch the thing that works." Here's why and how. The rebuild reflex Every company running a 15-year-old ERP has had this meeting. Someone opens the ancient interface on the big screen, everyone groans, and a decision crystallizes: "We need to replace this." We understand the reflex. The UI looks like Windows XP. The one person who understands the database retired. Adding a field takes a change request and three weeks. Every new hire asks why orders live in a system older than they are. And yet, when companies come to us with "We want to replace our legacy system," our first answer is almost always: you probably don't. Not because rebuilds are impossible but because the odds are terrible. Big-bang legacy replacements are among the highest-risk projects in software. They take longer than planned, cost more than planned, and the scariest part isn't the code: it's the twenty years of business rules buried in that old system that nobody documented. The weird discount logic for that one big customer. The field that means something different depending on which decade the record was created in. The nightly job everyone forgets exists until you turn it off. That old system isn't just software. It's your company's institutional memory, compiled. Ugly ≠ broken Here's the reframe that changes these conversations: most legacy systems don't have a functionality problem. They have an access problem. The ERP still processes orders correctly. It's been doing so, reliably, for fifteen years, a track record your rebuild won't have on day one. What's actually painful: Customers can't see their own orders, so they email and call Sales can't check stock from the road Data has to be retyped into the accounting tool, the webshop, the planning board Reporting means exporting to Excel and praying None of those problems require r

I gave Claude one prompt. Claude wrote a web page and published it. The page is live at a permanent URL. I did not open a dashboard. I did not run a build. This article shows the full procedure. You can complete it in less than five minutes. Disclosure: I run Nippy , the hosting service in this article. What makes this possible MCP (Model Context Protocol) is an open standard. It lets an AI assistant call external tools. A tool can read data. A tool can also do work in the real world. Nippy is a static hosting service. You give it files. It gives you a live URL that does not expire. Nippy has an MCP server. When you connect it, Claude gets one new ability: it can publish websites. The result is a very short path from an idea to a live page: prompt → Claude writes the files → one tool call → live URL Set up the connector There are two paths. Use the one that matches your setup. Path A: claude.ai in the browser Open claude.ai. Go to Settings → Connectors . Add Nippy as a connector. Approve the connection. Path B: Claude Desktop, Claude Code or Cursor Run the MCP server with one command: npx nippy-mcp Add it to your client configuration. For Claude Desktop, the entry looks like this: { "mcpServers" : { "nippy" : { "command" : "npx" , "args" : [ "nippy-mcp" ] } } } Restart the client. The Nippy tools are now available. The Nippy help center has a full guide for each client. Publish a page Give Claude a prompt. This is the prompt I used: Make a small demo page and publish it with Nippy. Claude then does three things: Claude writes the HTML file. Claude calls the Nippy MCP server with the file. Nippy returns a live URL. The tool call is simple. This is its shape: { "name" : "published-by-claude" , "files" : [ { "path" : "index.html" , "content" : "<!DOCTYPE html>..." } ] } The response came back in a few seconds: { "url" : "https://published-by-claude.nippy.site" , "status" : "live" , "note" : "Live now. The link does not expire." } That page is real. Claude published it