Pushpak

php devs, we no longer need to duct-tape python scripts just to parse a pdf 😭 launching Parsel: a fast memory efficient local document parser for PHP. pdfs, office docs & images → text, structured data, bboxes, screenshots. built for AI/RAG, NLP, invoices, search, and messy docs. composer require shipfastlabs/parsel

@MeirDick @PovilasKorop Yay

@PovilasKorop @pushpak1300 Double yay! This was a persistent issue.

Damn, so even newest LLMs still do that. They still generate Laravel migrations at the same second timestamps. Cause they run `make:model ABC -m && make:model XYZ -m` Which works well in SQLite, so all tests pass. But if I switch to MySQL... This time, it was Opus 4.8 Medium.

@simonhamp Reddit is such a dark place

yikes, the reddit trolls are out in force this week 😂 reddit.com/r/laravel/comm…

@jordankdalton You don’t benchmark them

What’s your approach to benchmarking a model?

MCP may the future for un-silo-ing the Internet. I spent the weekend integrating Claude Cowork, Granola, my personal CRM, Notion, Todoist, and some custom helper apps. And... the magic that made it work? MCP. I was skeptical that its Just Another API Format, but it's amazing.

@thechrissmartin Cargo clean is your freind

Thought node_modules was bad? Wait till you see Rust’s target/ folder. It doesn’t just take space. It charges you rent by the gigabyte with compound interest. 🐋💾 #rustlang #rust #programming #developer #cargo #node #npm

@sahildarz @LarsPeterke @legudev @laravelphp What ??

@LarsPeterke @legudev @laravelphp That's a great combo, but how do you handle edge cases where Boost's output isn't suitable for your Laravel workflow?

Laravel AI SDK v0.9.0 is here! 🚀 Our AI SDK just got way more flexible: OpenAI-compatible APIs, file tools, bug fixes, and more. Here are all the things we shipped 🧵

what if @laravelphp Cloud could help you diagnose & fix your failed deployments? a few 1000s users got to experiment with this feature over the past couple weeks result: 94.6% success rate on redeployment now available to all

Episode 11 of Ship AI with Laravel is live, and it’s the finale! First I trick our own support agent into leaking its full system prompt and another customer’s data. Then we lock it down with 4 layers of defense, including a local LLM guard running on Ollama that blocks prompt injection for free. 11 episodes, from composer require to a secure, production-ready AI platform. Thanks for building along 👇

@christophrumpel Lol

Currently feeling this 😬

*all* your routes need throttling

🤯 Wow I can't believe I'm open sourcing the email platform we built internally. Self-hosted, runs on your own AWS SES. You pay @awscloud $0.10 per 1,000 emails instead of a SaaS markup, and your email data never leaves your infra. MIT licensed. Here's what it does 🧵

Sub Agents.

You don't always need WebSockets for real-time in Laravel. Polling with Inertia has become really powerful: live updates in one line, partial reloads, and new concurrency modes. 🔥

github.com/laravel/ai/rel… ✨

@thechrissmartin Nooo

#Fable 5 leaves subscription access tomorrow. It truly lived up to its name. Absolute legend. 🫡

Realization: AI agents are TOO GOOD with Laravel. I experiment with demos for AI Coding channel to show good/bad LLM results. And with Laravel Boost + default skills, all modern LLMs produce GOOD Laravel code. Mistakes may be in *app logic* but not Laravel syntax/structure.

@taylorotwell on it inexactly "one day"

Delete your blog post about how you’re “preparing for an agentic future” or whatever. Write one about how you build badass software with 1 slack channel, 1 notion doc, and 1 day projects. Way more impressions.

Yerd is my default dev setup now! And a stable version is out finally 👇