Ez.-
banner
ecura.bsky.social
Ez.-
@ecura.bsky.social
Passionate for performance engineering.
Pinned
This is the era of disruption, start running!🏃‍♀️

It's a fascinating time when 2ppl from Barcelona can create tech beyond what all well financed vibecoding apps are able to provide you, yet.

Check out our new desplega.ai MCP!

#vibecoding #mcp #E2E #webapp
This year I learned the dutch opening, it was fun!

Thanks for the amazing recap @lichess.org!
December 18, 2025 at 9:49 AM
Have you tried antigravity.google?

I loved it until they started to use black hat techniques...
Antigravity lost me in 30 seconds, after an incredible first day.
I loved it so much, I closed Cursor immediately… blackhat manipulation techniques, & bad payment flows, lost me in seconds.
www.pleasedontdeploy.com
December 17, 2025 at 2:46 PM
I've been having so much fun going through PW issues this last week, that I created this game:

www.desplega.ai/tools/flaky-...

Suggestions welcome!
desplega.ai - AI-Powered E2E Testing | Stop Trading Quality for Speed
Transform your QA with AI-powered end-to-end testing. Generate, maintain, and run tests automatically. Achieve faster releases without compromising quality.
www.desplega.ai
December 5, 2025 at 12:21 PM
Just shipped something kinda interesting for #lovable

It runs e2e tests from their site, without you doing anything.

Looking for those first people that will roast it properly. We are adding 500 extra tokens for the first few.

youtu.be/Z5bvjVZZpF8
chromewebstore.google.com/detail/vibe-...
Vibe QA: AI-powered Testing for Lovable.dev, Bolt & More - Chrome Web Store
Add-On for Lovable.dev, Bolt and other vibe coding tools to test your apps with desplega.ai QA engine.
chromewebstore.google.com
November 30, 2025 at 2:38 PM
Anyone playing around with labs.google/portraits/gallery?

I've been having basic conversations with Nir Eyal & Kim Scott, but I'm sure there's more to do there.
Conversations with people who inspire, powered by AI
labs.google
November 26, 2025 at 2:33 PM
“I believe the hard part of building software to be the specification, design, and testing of this conceptual construct, not the labor of representing it and testing the fidelity of the representation.”

Change my mind.
November 10, 2025 at 5:53 PM
SDKs for LLMs? Frameworks meant to never be human-read? Who's cranking these out?

I'm looking for OSS optimized to be used by LLM generated code.

If it's not you, who is it?

How can I never look at code again? 😏
November 6, 2025 at 12:22 PM
TIL: having a child forces you to connect with society in a different way.

It's like moving from IC to Manager... I think I can finally understand why respect for elders was always so important, those are the true Executives.

Promoted myself to clueless middle management. Send help (and snacks).
November 4, 2025 at 12:16 PM
Waymo just learned how to get away with a traffic ticket!

The piece that tops it all is that the officer behaved in a more structured way than the machine, ie. the self-driving car allegedly did a wrongful U-turn, maybe to save time?
October 28, 2025 at 5:38 PM
"Just do it!" doesn't quite work for ChatGPT Atlas.

If you are like me, you are going to use your weekend to tinker with Atlas, before you do so, check out two things:

+ Blind Goal-Directedness problem. (arxiv.org/pdf/2510.01670)
+ Agentic AI’s OODA Loop Problem.(www.schneier.com/blog/archive...)
October 24, 2025 at 5:12 PM
Back in 2021 I forbade the term 'Tech debt' in my teams.

That was the only way I found to be explicit as an organization: every effort we made needed to have a business reason behind it. It's never about the debt, it's always about how costly it is to cancel it.

www.mckinsey.com/capabilities...
www.mckinsey.com
October 22, 2025 at 1:12 PM
"We are not longer in this era where good enough is fine; good enough is mediocre. You need to get to great, probably excellent."

If you haven't noticed this, it's probably because you are still wondering how to accelerate. Soon afterwards you are going to realize something...
October 21, 2025 at 4:38 PM
How many do you think we were?
October 19, 2025 at 9:12 PM
Everything is fake! What are your trusted sources for journalism?

With sora-2, nanobanana, and LLMs flooding our webpages, I'm starting to believe well curated journalism [video/photo/audio/text] would be the next big thing.
October 17, 2025 at 5:12 PM
I'm not slacking, claude is #vibing!

That's how it felt during the first 15 yrs of a SWE career. The key: learn how to multitask at a different level. And Yes! Everyone will be a (vibe)coder of sorts.

I know `I'm absolutely right`.
xkcd.com/303/
October 15, 2025 at 1:12 PM
#9 in hacker news last Friday.

That's how interesting www.npmjs.com/package/@des... is.

The fun never stops.
October 14, 2025 at 4:38 PM
This is the era of disruption, start running!🏃‍♀️

It's a fascinating time when 2ppl from Barcelona can create tech beyond what all well financed vibecoding apps are able to provide you, yet.

Check out our new desplega.ai MCP!

#vibecoding #mcp #E2E #webapp
October 6, 2025 at 8:39 AM
$1.52 trillion - cost of technical debt in the US alone in 2022, back then
87% of CIOs said they were slowed down by that debt.

Let that sink in.

CISQ'22 report: buff.ly/JijHd7u
The future of your code is no-code.
Is your codebase being reduced every day?
open.substack.com
September 30, 2025 at 4:38 PM
"What does a high-level language accomplish? It frees a program from much of its accidental complexity."

Vibe-coding is the new high-level coding language, well beyond what Frederic Brooks and Parnas envisioned in the 80s.
September 26, 2025 at 5:12 PM
"If you think it needs to happen quickly, there's a good chance you are actually not really committed to it."
buff.ly/NilOUC9
September 19, 2025 at 5:12 PM
Don't hire your first VP of Eng!

Once again, I find myself having the exact same convo when a startup hits ~15 SWEs. The default advice they received is to hire a VP... don't!
September 17, 2025 at 1:12 PM
QA is the next critical challenge, and Amjad Masad (Replit CEO) just showed it 🔥.

Replit launched Automated Testing last week, and we decided to test it out.

We asked it to create a todo app.
We waited...
We waited...
1hr later, we had a barebones app, and +90% of our credits were spent on...
September 16, 2025 at 4:38 PM
"I’ve often heard, with decent reason, an LLM compared to a junior colleague. But I find LLMs are quite happy to say “all tests green”, yet when I run them, there are failures. If that was a junior engineer’s behavior, how long would it be before H.R. was involved?"
September 12, 2025 at 5:12 PM
One of my top concerns nowadays: How do I make it to 2000 ELO in chess.com?

I want to be able to go back to the club scene and feel like the old days... that I have a chance to win against a kid.

To be fair, I used to be the kid...
Chess.com - Play Chess Online - Free Games
Play chess online for free on Chess.com with over 200 million members from around the world. Have fun playing with friends or challenging the computer!
chess.com
September 11, 2025 at 5:32 PM
The AI latency tax is real, and everyone is paying it!

Have you found yourself going back to Google search when asking simple questions? I do, because some things should take less than 5 seconds.
September 9, 2025 at 4:38 PM