Configuring Bref to deploy Laravel storage S3 buckets got much simpler. S3 is simple, but large file uploads straight to S3 requires: - enabling S3 ACLs (deprecated feature, but Laravel still uses it) - enabling cors The new settings make it much simpler, here's a before/after.

I wrote a detailed blog post: mnapoli.fr/running-gith...

I moved my GitHub Actions to runners on my Mac Mini, my CI just got muuuch faster. That works for me as a solo dev, I have 4 runners (could probably add 4 more). For teams I wouldn't recommend it tbh, I'd look into Depot CI or Runs-On, that'd make much more sense.

good thing I posted, I'm changing to this:

Instead of bloating environment variables, I've created an `envMap()` helper in Laravel config. This is for values that : - are not secrets - have hardcoded values depending on the environment Thoughts?

When you create a private network (VPC) with Bref Cloud, you can now copy the IP addresses used for outgoing traffic. e.g. if you need to configure an API to only accept traffic from your app's IP address

with all the supply chain attacks, I'm pushing code to GitHub the secure way

Released exspec 0.3: - switched from Playwright MCP to CLI -> faster, less tokens - claude runs the specs as test, but sometimes the spec is not clear -> claude now suggests recommendations for confusing tests - more efficient, more logs, more reliable github.com/mnapoli/exs...

🤣 Merci @mnapoli.bsky.social 🤩

Les attaques de supply chain évoluent, méfiez vous

I think I've got something wrong somewhere, but I can't pinpoint it for sure 🤔

[experimental] trying something new with testing, looking for feedback Problem: coding accelerates, I can't keep up with tests. Either I review everything and we're not moving faster, either I don't and don't trust them Idea: executable specs

If you're getting old like me, writing long prompts in a terminal is annoying: can't use the mouse, select text, use normal shortcuts… I built a native Claude Code prompt editor: hit Ctrl+G, edit the prompt in an overlay window and hit save github.com/mnapoli/Pro...

New Bref documentation page: Cold Starts bref.sh/docs/enviro... (should have done it years ago but better late than never)

that moment when a new project hits prod ✨ (also Laravel Octane on bref.sh those response times are awesome)

The "push -> PR review -> fix -> PR review -> fix" loop works, but it's slow. I built a PhpStorm plugin that: - runs Claude + Codex review locally - they create inline comments *in PhpStorm* - you review/edit/dismiss comments - click to copy a prompt for Claude/Codex to fix them all

I run `/address-pr-review` all the time It reads PR comments and addresses them. I use it in a loop (push -> fix -> push -> fix): the skill ignores resolved comments and replies to open threads ("fixed" or "won't fix because …") I shared it on GitHub: github.com/mnapoli/ski...

Taylor Otwell casually shipping a broken app online. The app triggers an alert, which triggers opencode which writes a fix, pushes a PR, then OpenClaw calls Taylor BY PHONE on stage, who responds « merge it », and the fix is merged and deployed. All in 2 minutes, live on stage 😂

Preparing for LaraconEU I have now 2 floppy disk readers, I can create twice as much 💪 Come find me if you want one!

Bref 3 is released 🎉 Since its creation, Bref has served 1.2 trillion requests and jobs! ▸ simpler & lighter runtimes ▸ PHP 8.5 ▸ improved observability (logs & tracing) ▸ simpler Laravel onboarding ▸ postgres & redis built-in and more ⤵ bref.sh/news/03-bre...

Working on Bref v3 release for tomorrow 🤞

Most emails I receive about open source are unsolicited feature requests, but not all emails 😌

Improved Bref docs on a quick win: cleanup the AWS SDK before deployment bref.sh/docs/deploy... (btw could also be useful if you deploy your app on a server)

I don't understand why PHP got typed constants. Constants never change, their type is exactly what the constant is.

New UI to configure private networks in bref.sh/cloud

Replace Claude rules with automated checks, run this now this is so good: ``` Read all CLAUDE.md files in the project, along with the existing tool configurations (eslint, phpstan, pint, package.json scripts, git hooks, and everything else relevant).

I used the new Laravel AI SDK and made a dumb mistake: it's really for implementing agents, not prompting models. I had a task with a single request to send to Gemini (transcribe this image into text) and using Prism directly made much more sense in the end. TIL!

Trying out the new "agent teams" in Claude Code. I had very low expectations (I plan small tasks, "no need for a team"), but I am blown away so far.

I tried the Codex App but didn't have a great experience. I tried both GPT 5.2 and GPT 5.2 Codex (high or ultrahigh), they are incredibly slow. (that was just before the release of GPT 5.3, dunno if that's different now)

Claude keeps writing transactions wrong, I wrote this skill in case you're interested gist.github.com/mnapoli/0e8... (also if you spot something wrong let me know 🙏)

In Laravel apps that heavily use actions (e.g. most entities are created via actions), do you use factories at all? E.g. `CreateUser` action does advanced stuff. Then `UserFactory` is not aligned and tests are all wrong. Factories don't work in this case, right?

I was asked about Claude Code with Beads, might as well post here too. I used Beads for a while: - 1 terminal to create issues, or even epics for large tasks - 1+ terminals to implement them Now I moved to:

ok Claude Code has the cool factor but going all in on terminal UI is such a mistake IMO… OpenAI launched a Codex app. Like Claude Code but as a macOS app. The UX looks incredible! Just watch the video! Which OpenAI model to use instead of Opus 4.5? openai.com/index/intro...

My current AI coding workflow: 1. plan with Claude Code -> create GitHub issue 2. new session: `/work #122` (+ I review the code locally) 3. `/pr` creates a PR 4. Claude reviews in GitHub Action and adds inline comments 5. I review the diff + comments (mark some as resolved)

I tested StageHand for writing high level tests. Interesting but I think we need even higher level tests today. I don't think it's the right tool for this. docs.stagehand.dev/v3/first-st...

Prototyped a MCP server that lets agents use high-level tools instead of grep/edit: - rename class/… - find definition - find usage Used phpactor under the hood. It works, but phpactor is nowhere as good as PhpStorm or PhpStan to understand codebases. Back to square one.

I'm not the kind to spend days tweaking the terminal UI and themes but I played with it this weekend and that font and theme in claude 🤌🤌 Font is "Space Mono for Powerline" TIL about Powerline too (theme is Nord Aurora) And the Claude status line: github.com/sirmalloc/c...

I create an app to follow Claude Pro or Max subscription usage Other apps I tested either report $$ usage that doesn't match the fixed cost of the Pro/Max plan, or unrealistic percentage. This app gets the data straight from your Claude account page. github.com/mnapoli/cla...

New Laravel package: add `temporaryUploadUrl()` support to the local storage If you use S3 in production, you probably use `temporaryUploadUrl()`. That package adds support for that locally so that you don't have to adapt your code. github.com/mnapoli/lar...

I created the Serverless Visually Explained course 6 years ago. It's now free! If you want concrete JS and PHP examples, with live animations, and no-bullshit explanations on what serverless is, you can click and read. serverless-visually-explained.com/

Tested OpenCode vs Claude Code: Claude wrote better code, 2× faster, 30% cheaper I asked both to implement the same plan: - Claude: 14 min, ~191k tokens - OpenCode: 27 min, ~278k tokens Quality was close (and good), Claude slightly better overall.

Trying OpenCode: - incompatible with Warp, had to switch to Ghostty 👎 - migration from Claude Code sucks, nothing automatic - the UX is great 👍 (lost some Warp shortcuts though) - no Claude Code web 👎 Total migration time: ~3 hours 😢 Now testing "is it actually better than CC?"

In 2014 I was all about dependency injection now I'm arguing *against* DI in many scenarios Would this make for a good talk? How 🌶️ should it be?

I just discovered `modeltyper` to automatically generate TypeScript types from Eloquent models, works great! github.com/fumeapp/mod...

Worked on a Claude skill so that one could ask "set up Bref" in any PHP project. It _looked_ awesome: Claude would analyze the project, ask questions, etc. But it's a fail. As usual, Claude just invents options, doesn't follow the script reliably, uses its outdated knowledge…

Application diagrams are live in Bref Cloud ✨ They'll help understand how serverless applications run in AWS: - CloudFront CDN - S3 for assets - PHP runs on Lambda - SQS queues… Metrics are live: requests, latency, scaling, errors… beautiful design by @nunowar!

GitHub notifications now end up on my desk

Claude Code users (especially Laravel/Symfony): you REALLY want to re-install Claude Latest Claude versions will load your `.env` (including secrets!) into Claude Code. Claude then runs your tests with local config instead of testing config!

At least Claude is honest I guess

I often get asked about the differences between serverless and containers/servers I updated Bref docs to be much more upfront about what to expect to migrate to serverless. Am I missing something? What's unclear? You can comment the PR here: github.com/brefphp/bre... Docs here: bref.sh/docs