标签：#t

Remember how I promised you (or rather myself) two weeks ago that from now on I'd only write light,...

I remember the exact moment my first AI side project died. It was 3 AM, I had just spent two full weeks building an elaborate RAG pipeline with vector databases, custom embeddings, and a fine-tuned model—all for a tool that would "revolutionize how developers read documentation." I hadn't written a single line of user-facing code. I hadn't even validated if anyone wanted it. And when I finally deployed it to a hobby server, the cost of hosting the model alone was $200/month. I killed the project before anyone ever visited the URL. That was three months ago. Since then, I've shipped three AI side projects that actually have users. Not millions—but real people who use them daily. Two of them even cover their own hosting costs now. The difference? I stopped trying to build the perfect AI infrastructure and started shipping the stupidest thing that could work. Here's what I learned from those three MVPs, and how you can break out of the "AI side project graveyard" too. The Trap: Thinking You Need to Build Everything The biggest lie in the AI side project space is that you need to own the stack. Every tutorial screams "self-host Llama 3," "set up your own vector database," "build a custom agent framework." That's great for learning, but it's death for shipping. For my second project—a tool that automatically generates commit messages from diffs—I spent exactly one evening. I used the OpenAI API directly, with no caching, no streaming, no error handling. Here's the core of it: import openai import subprocess def get_diff (): result = subprocess . run ([ " git " , " diff " , " --cached " ], capture_output = True , text = True ) return result . stdout def generate_commit_message ( diff ): response = openai . chat . completions . create ( model = " gpt-3.5-turbo " , messages = [ { " role " : " system " , " content " : " Write a concise git commit message summarizing the changes. " }, { " role " : " user " , " content " : diff } ] ) return response . choices [ 0 ]. message .

It's one of the best times to snag yourself a Dyson device, whether it's a vacuum or a beauty tool.

TL;DR for vibe coders: Shipped an app with Cursor, Claude Code, or v0 and got a scary Vercel or Neon bill? You probably don't need a bigger plan. You need a few fixes. Not technical? Copy the agent-rules folder from the companion repo into your project and tell your AI editor "apply these cost rules." That's the whole job. You don't need Vercel Pro. You don't need to "Launch" on Neon. A lot of the time you need the opposite of an upgrade. You need your stack to go to sleep. Here's the moment that started this. A Neon bill landed, across three small Next.js apps running on Vercel with a Neon Postgres database, and the breakdown was lopsided in a way that gave the whole game away. $32.65 of it was compute. Storage was 5 cents. History and data transfer were zero. The meter said 308 hours of compute time in about three weeks, which is a database awake for roughly fifteen hours a day, every day. The instinct when a bill climbs is "I must be getting traffic, time for a bigger plan." So before doing that, I opened Google Analytics. Zero users in the last 30 minutes. Meanwhile the database compute was pinned active, and had been more or less around the clock. So this was never a storage problem or a traffic problem. Almost the entire bill was one thing: paying for a database to stay awake doing nothing. Both of those things were true at the same time, and here's the part I'll say up front so nobody has to guess: I didn't hand-write the mistakes that caused this. AI coding agents did. I described features, the agent shipped working code, and that code carried a handful of patterns that quietly defeat scale-to-zero. That's not a confession of bad engineering. It's just what building looks like in 2026. Most of us are vibe-coding onto serverless and cloud infra now, and the agent optimizes for "make the feature work," not "keep the database asleep when no one's around." It has no idea your compute bills by the hour it stays awake, so it has no reason to care. That's the whole

Every developer using Cursor , Claude Code , Windsurf , or GitHub Copilot knows this exact frustration: You are building a cutting-edge Angular 22 application. You ask your AI coding assistant to spin up a dynamic form, a lazy-loaded list, or an asynchronous data card. Instead of leveraging modern fine-grained reactive Signals, optimized native block control flows, or proper SSR hydration hooks, the AI drops an unoptimized pile of legacy tech debt full of NgModules , *ngIf , *ngFor , and raw RxJS BehaviorSubjects . The LLM Training Paradox Why does this happen? Large Language Models are trained on historical code datasets. Statistically, more than 90% of the public Angular repositories and StackOverflow threads on the internet represent older paradigms. Left to their own devices, agents default to the statistical average of their training data. They literally default to the past. The Fix: angular22-agent-skills To solve this, I built a public, open-source repository of custom instruction bundles and system guardrails leveraging the new skills.sh tool standard. By injecting this verified context directly into your development environment, you force your local AI agents to bypass their training averages and write pristine, optimized, modern Angular 22 syntax every single time. 👉 Check out the repo here: https://github.com/PavanAnguluri/angular22-agent-skills 🔍 The Difference: Before vs. After To understand why these guardrails are necessary, look at what an AI agent writes out of the box versus what it writes once you apply the angular22-agent-skills harness. 🚫 What AI Agents Generate by Default (Legacy) // The AI falls back to old decorators and heavy RxJS boilerplate for standard state import { Component , Input , OnInit } from ' @angular/core ' ; import { BehaviorSubject } from ' rxjs ' ; @ Component ({ selector : ' app-user-profile ' , template : ` <div *ngIf="visible"> <h3>{{ firstName }} {{ lastName }}</h3> <div *ngFor="let item of items"> {{ item.name }} </div>

The Big Lie in Mobile Privacy (And How We Fixed It) If you have an Android phone, you've seen the pop-up banners asking for your permission to track your data. You click "Reject All," assuming the app stops tracking you. Here is the dirty secret of the mobile app industry: It usually doesn't stop them. 🔍 The Pipeline Problem Traditional privacy tools are built like internet filters. When an app tries to send your data across the web to a data company, the privacy tool tries to block that specific web traffic. There is a massive flaw in this approach: The moment you open an app, hidden tracking packages (called SDKs) wake up instantly. They immediately copy your phone's ID, your location, and your usage habits into their internal memory. Even if a network filter blocks them from sending it right now, your data is already collected. The trackers just wait until the filter drops, or they find a workaround to leak it out later. You are forced to blindly trust that these third-party trackers will behave themselves. Spoiler alert: they don't. 🔒 CookiePrime: Locking the Front Door At CookiePrime, we got tired of the "illusion" of privacy. Founded by privacy industry veterans who witnessed how easily corporate trackers bypass traditional regulations, we decided to build a true privacy enforcement ecosystem . Instead of trying to catch your data as it flies out over the internet, our Android software stops trackers from waking up in the first place. Think of trackers like uninvited snoops at a party: Traditional tools try to grab the snoop's notebook after they've walked around your house and written down your secrets. CookiePrime locks the front door so the snoop never steps inside. The moment a CookiePrime-protected app starts, our engine runs a lightning-fast sweep — taking just 93 milliseconds — to identify every tracking script hidden inside the app. If a user says "No Tracking," CookiePrime instantly freezes those specific trackers on the spot. They can't collect data,

Walmart-backed Flipkart has crossed 1,000 micro-fulfillment centers as Amazon accelerates its own quick-commerce push in India.

If you drive the WordPress admin via Playwright for long enough, one day a screen you've never seen before will appear after login , and everything downstream stops working. Is admin@example.com still the correct admin email address? [ Yes, the email is correct ] [ Change the address ] That's WordPress's admin email confirmation screen . Roughly every six months, after the admin user logs in, this confirmation screen gets injected — standard behavior since WP 4.9. A human just clicks once. An automation script can't see it without explicit handling. Why automation gets stuck A straightforward Playwright login looks like: page . fill ( ' #user_login ' , user ) page . fill ( ' #user_pass ' , pwd ) page . click ( ' input[type= " submit " ] ' ) page . wait_for_load_state ( ' domcontentloaded ' ) # Assumes we're on the dashboard page . goto ( ' /wp-admin/plugins.php ' ) But on a "confirmation day," the URL right after wait_for_load_state is something like /wp-admin/profile.php?...action=confirm_admin_email... — the confirmation screen. You thought you were navigating to the plugins page, but the DOM you expected isn't there. Subsequent selectors fail, and everything downstream cascades into failure. A specific selector identifies the screen WordPress's confirmation screen has a uniquely-named submit button: <input type= "submit" name= "correct-admin-email" value= "Yes, the email is correct" /> If input[name="correct-admin-email"] exists on the page, you're on the confirmation screen. The same selector serves as both the detection signal and the click target , so handling is only a few lines: admin_email_confirm = page . locator ( ' input[type= " submit " ][name= " correct-admin-email " ] ' ) if admin_email_confirm . count () > 0 : logger . info ( " Confirmation screen detected — clicking ' email is correct '" ) admin_email_confirm . first . click () page . wait_for_load_state ( ' domcontentloaded ' , timeout = 30000 ) Insert this after the post-login wait_for_load_state

The Surface Laptop is down to $835 for Prime Day—a killer discount on one of my favorite laptops.

Every identity migration guide eventually reaches the same paragraph, and it's always a little apologetic: "users will need to reset their passwords." It gets treated like a law of nature. It isn't. It's a choice, usually forced by a tool that didn't want to do the harder thing. The harder thing is verifying your users' existing password hashes in place, so they sign in after the move with exactly the credentials they had before and never notice anything happened. Whether you can do it comes down to one question: can you get the old hashes, and can the new system verify them? Password hashes are more portable than people think A password hash isn't a secret algorithm. bcrypt is bcrypt. A bcrypt hash carries its own cost factor and salt inside the string, so anything that implements bcrypt can verify a hash any other bcrypt system produced. The same is true of the PBKDF2 format ASP.NET Identity uses: documented, versioned, self-describing. If you know what you're holding, you can check a password against it without ever knowing the password. So a migration that preserves logins doesn't need the plaintext (nobody has it) and doesn't need to re-hash everyone up front. It needs to obtain the stored hashes and verify against them on sign-in, upgrading each one to its own format quietly the first time a user logs in. That last part is lazy migration: carry the old hash, verify it once, replace it transparently. Over a few weeks of normal logins your user table re-hashes itself and the legacy formats age out, with zero resets and zero support tickets. The dual-path bit The wrinkle is that different sources hand you different formats, and a good importer verifies both: From self-hosted Duende / ASP.NET Identity: the V3 PBKDF2 hashes (and any legacy bcrypt) verify natively and rehash on first sign-in. This is the easy case, because it's the same scheme the destination already uses. Most teams are surprised it's that clean. From Auth0: bcrypt hashes verify verbatim. The catch

The first time I wired Dependabot up to a private package registry, it took three meetings and a calendar reminder set six months ahead. The reminder was for the personal access token I had to mint to do it, the one I would have to rotate by hand before it expired, living in a config that drifted somewhere between repo settings, a .github file, and an internal wiki page nobody had touched in a year. On June 23 GitHub posted a small changelog item that quietly retires that whole ritual: Dependabot can now read your private GitHub Packages and GHCR registries through its own GITHUB_TOKEN , no PAT required. If you have ever had to explain to a security review why your bot account holds a token your team owns but a person minted, you already know why this lands. The PAT we all had and nobody loved Until this week, Dependabot's path to a private GitHub-hosted package was the same path it took years ago. A human minted a personal access token, scoped wide enough to read the registry, stored it as a repo or org secret, and plugged it into Dependabot so the bot could authenticate when it tried to resolve dependencies. That token expired on a human schedule. It belonged to whoever happened to set Dependabot up, which is rarely the person who still owns the repo a year later. You can feel how that ends. Tokens that quietly expire and break a Dependabot PR the day someone needs the patch. Tokens that follow an offboarded engineer out the door. Tokens with scopes wider than the job actually requires, because nobody wanted to mint a fine-grained one per registry path. None of those failures are catastrophic on their own. All of them are the kind of papercut a platform team ends up owning. What the GITHUB_TOKEN does here The new behavior is straightforward. Dependabot's job-scoped GITHUB_TOKEN can now request a packages: read permission. When the bot pulls from a hostname under *.pkg.github.com or ghcr.io , it sends that token instead of asking for a PAT. The same token that alre

Top 7 Featured DEV Posts of the Week Cyberpunk cat RPGs and robot personalities Jess Lee Jess Lee Jess Lee Follow for The DEV Team Jun 23 Top 7 Featured DEV Posts of the Week # discuss # top7 34 reactions 7 comments 2 min read

I'm Dr. Mohammad Reza Beheshti, Founder of CyberSiARA. I hold a PhD in Electronic Engineering and Artificial Intelligence and have over 15 years of experience in cybersecurity research and innovation. My passion has always been solving complex security challenges through technology. This journey led me to found CyberSiARA, where we're developing AI-powered bot protection and human verification solutions to help organizations defend against increasingly sophisticated cyber threats. I enjoy combining academic research with practical engineering to create technologies that are both innovative and effective in the real world. Through this blog, I share insights from my research, product development, and experiences building a cybersecurity company, with the aim of helping developers and security professionals stay ahead of emerging threats. I'm always keen to learn, collaborate, and contribute to the global developer and cybersecurity communities.

The boat-agent stack here runs on a prime directive: if there's something usable out there, improve it; build our own only as a last resort. So when we needed a SignalK MCP server, the honest first move wasn't to write one — it was to evaluate the one that already exists. VesselSense/signalk-mcp-server (TypeScript, MIT) is good work. It exposes SignalK to an agent through a single execute_code tool: the model writes JavaScript, the server runs it in a sandboxed V8 isolate ( isolated-vm ), and only the result comes back. Its README claims a 90–96% token reduction versus traditional named MCP tools — 2,000 tokens down to 120 for a vessel-state query, 13,000 down to 300 for a multi-call workflow. Those numbers are plausible, and they line up with the broader industry result that code execution beats tool-calling on token efficiency for complex multi-step work. We read it, ran the numbers against our own agent, and kept our discrete-named-tool signalk-mcp anyway — then harvested three of VesselSense's ideas into our roadmap. This post is that evaluation: the two philosophies, why the obvious-sounding win doesn't bind for a voice-first agent, and a decision framework you can reuse before you adopt-or-build your own MCP server. This is a design-reasoning post, not a debugging saga, but it maps to the same arc: a question, the dead-end that looks like an obvious yes, and the call that actually held. The question Two SignalK MCP servers, two genuinely different designs: VesselSense/signalk-mcp-server sailingnaturali/signalk-mcp ───────────────────────────── ─────────────────────────── one tool: execute_code discrete named tools: → agent writes JavaScript read_sensor(path) → runs in a V8 isolate battery_state(bank) → queries SignalK, returns depth_state() only the result get_route() get_local_time() TypeScript / Node + isolated-vm list_paths(prefix) claims 90–96% fewer tokens get_active_alarms() Python, end-to-end The adopt-vs-keep question: does the token-efficiency win bin

We run a prime directive on this stack: if a usable tool already exists, improve it; build our own only as a last resort, and when you keep your own, record why each alternative failed. This post is that audit for weather-mcp — a marine-weather MCP server — against the weather-MCP ecosystem, and the one capability change that fell out of it. The short version: three perfectly good weather MCP servers exist, and none of them does the thing a navigator actually needs. The reasons generalize to any "adopt an MCP server or keep your own" call, so the audit is the post. Then the fix — parsing a second NDBC file format to split swell from wind waves — is small enough to paste in full, and it surfaced data the standard file had thrown away. The problem, as you'd search it You want an agent to answer "what are the seas doing where we are?" and you go looking for a marine weather MCP. You find a few. Each one returns a forecast . None of them returns what a buoy 12 nautical miles away is measuring right now . That gap — forecast vs. observed — is the entire job, and it's the one thing the ecosystem skips. Here's what's on the shelf, and what each one is missing for marine use. The candidates Three real servers, all worth your time for what they're built for: cmer81/open-meteo-mcp ~13 tools, raw Open-Meteo JSON straight through weather-mcp/weather-mcp ~12 tools, own format, global; marine = Open-Meteo RyanCardin15/NOAA-Tides... CO-OPS stations: water levels + currents, not buoys And ours: sailingnaturali/weather-mcp 4 tools, Python, 2 runtime deps (httpx + mcp) get_marine_forecast Open-Meteo wind/swell/wind-wave/seas/pressure get_marine_forecast_premium Stormglass blend — 10 tokens/UTC-day, cache hits free get_nearest_buoy_observations NDBC observed wind + waves by lat/lon, with bearing + age get_stormglass_quota_status token-ledger read, no network Mapped against what a navigator needs: Capability ours open-meteo-mcp weather-mcp/weather-mcp NOAA-Tides Open-Meteo marine (swel

The all-cash deal gives MoEngage access to technology that assigns AI agents to individual customers.

A lot of American bathrooms don't have outlets near the toilet. Tushy Classic 3.0 and Wave bidets, on sale for Prime Day, solve the problem.

This is the third piece in a sequence. The first asked whether Sloan had flagged anyone else — it...

HiSense's new CanvasTV line is a better buy than the Samsung Frame, even at full price. And it's not full price right now.

The Nex Playground is the family-centric, Kinect-like game console that made one Verge editor’s kids laugh, cry, and ask for more playtime, even when they were sick. The motion-based game play isn’t perfect, but it has won over plenty of parents — and it’s on sale for $239 at Amazon, down from its current MSRP […]