TheProdSDE

@towards_AI @claudeai Because terminal looks cool? Jk terminal way is faster for most of coder and developers

@claudeai why are we going back to terminal?

New in Claude Code: agent view. One list of all your sessions, available today as a research preview.

@Priya_Upadhyay_ Just because it looks cool to me. Satisfaction is the key. Even with zero users it make me happy it worked

Why do we spend weeks building a scalable backend for a product that has zero users.

Check out A2A protocol with all details and how to use in prod at : pub.towardsai.net/your-ai-agents… Part 2 will have codes and part 3 will be deployment at azure

@dev_maims W

The same generation that once rejected cookies is now giving AI full access to their files, desktops, and even financial accounts.

@Priya_Upadhyay_ Well blaming the AI is not right if you don't know anything. obviously AI also make pricing based on the how dummy you are.

Everyone is a vibe coder until the api bill hits four figures in a single night.

@mattpocockuk @adamwathan Kinda boring to work on but yes will work if. pays well No doubt in that

@adamwathan Yes, because taste is hard to replicate, but copy + tweak beyond recognition is easy

This whole “get the DESIGN.md for any website and make your own website look like a shitty version of someone else’s website” thing is depressing 🥴 some of you actually want that?

@teej_dv make sense he was checking if you got any brain left or he can charge you more when token limit exhaust.

also - it didn't fix the problem

5.5 just edited a node_module file directly.... what?!

@Im_IrushiK Stop posting on X at least the picture and do digging yourself.

Just installed Linux for the first time. What should I do next ?

@ThePrimeagen kinda same for all run to new one rather than fixing the core problem0!

Current meta

@devXritesh keep one and return false for unverified Simple with the basic concept

A junior dev designed these two endpoints.... GET /api/v1/followers/verified & GET /api/v1/followers/unverified What’s the flaw in this design, and how would you improve it using the correct parameter approach?

@CodeWithAmann everything has it's flaws and that's beauty decision should come from use case not rigidness

Which backend stack has the strongest future?

Guide How to do vibe coding correctly to ship to prod: No Lies medium.com/towards-artifi… #AI #Software #VibeCoding

@ThePrimeagen They forgot whose accounts they are phishing:)

People really out there trying to Phish

#AITools #SoftwareEngineering #DeveloperProductivity Link: blog.stackademic.com/stop-using-you…

My premium AI credits used to vanish by Wednesday. Every week. Not because I was doing anything wrong. Because I was using my most powerful model for everything — planning, architecture, and implementation in one session. Here's the 5-step system that fixed it 🧵

#SoftwareEngineering #AITools #CodeReview

My teammate asked: "Did you actually read PR #117 before approving it?" I had. With AI. The code was fine. He wasn't asking about the code. He'd spent 3 weeks on a new pattern. That PR was his quiet test to see if anyone noticed. I didn't. I wrote about what happened next 🧵

Meet Kimi K2.6: Advancing Open-Source Coding 🔹Open-source SOTA on HLE w/ tools (54.0), SWE-Bench Pro (58.6), SWE-bench Multilingual (76.7), BrowseComp (83.2), Toolathlon (50.0), Charxiv w/ python(86.7), Math Vision w/ python (93.2) What's new: 🔹Long-horizon coding - 4,000+ tool calls, over 12 hours of continuous execution, with generalization across languages (Rust, Go, Python) and tasks (frontend, devops, perf optimization). 🔹Motion-rich frontend - Videos in hero sections, WebGL shaders, GSAP + Framer Motion, Three.js 3D. 🔹Agent Swarms, elevated - 300 parallel sub-agents × 4,000 steps per run (up from K2.5's 100 / 1,500). One prompt, 100+ files. 🔹Proactive Agents - K2.6 model powers OpenClaw, Hermes Agent, etc for 24/7 autonomous ops. 🔹Claw Groups (research preview) - bring your own agents, command your friends', bots & humans in the loop. - K2.6 is now live on kimi.com in chat mode and agent mode. For production-grade coding, pair K2.6 with Kimi Code: kimi.com/code - 🔗 API: platform.moonshot.ai 🔗 Tech blog: kimi.com/blog/kimi-k2-6 🔗 Weights & code: huggingface.co/moonshotai/Kim…

Full story + the local HTTPS setup failure that got me 👇 blog.stackademic.com/stop-using-you… #AIEngineering #SoftwareDevelopment #LLM #AIWorkflow #WebDev #Programming

My 3-tier workflow: 1️⃣ GPT-4o mini → write the full plan MD 2️⃣ You → review + fill gaps 3️⃣ Sonnet/Opus → AUDIT the plan only 4️⃣ You → verify 5️⃣ Right model → implement 1 criterion at a time Went from 100 credits in 2 days → 30-40 all month.

I wired up an SSO integration using AI. Before writing a single line of code, the AI audit found 12 security bugs already in the codebase: • Tokens in redirect URLs • localStorage storing JWTs • JWT decoded without signature check