yessss

I tried Fable extensively, I am blown away. Daily tasks are handled well, better than Opus but nothing too crazy. Complex tasks, risky features, etc -> Fable is insanely good. It codes better than

3 nice Bref updates: - zero-downtime deploys for website assets - public storage on S3 made simple - osls 4.0: resolved deprecations, hardened, supports SSO

Best job in the world… I hope the reports of its death are greatly exaggerated

I'm starting to really like Cursor automations: - CI broken on main -> create PR - Sentry error -> create PR - Review 1 random file every night -> post bugs found in Slack - Watch for new PHP releases (for Bref upgrades) etc.

Introducing VibePHP, a next-gen PHP runtime that runs on vibes and supports generics, async/await and more 🌀 github.com/mnapoli/vib... 🌀

Graphed my AI spend since 2024 lately it's mostly Cursor (main) + Claude, with a bit of OpenAI for ChatGPT and Gemini for specialized OCR

I rewrote externals.io to Laravel! (runs with Octane on Bref) 50% of requests under 5ms, p90 is 40ms 🚀 2.5M req/mo, costs: ~$2.5/mo + $11 for the database, staging costs $0 Open source on GitHub: github.com/mnapoli/ext...

Pest browser tests are way more insane than I thought. Classic browser tests hit a real server. Instead, Pest starts an HTTP server on the fly (random port) in the _same_ process as the test, and points chrome headless to it - no server to start

Conferences: steal this idea from Laravel Live Japan 🙏 no lunch buffet, lunch is outside with optional "lunch pairing" (groups of 4 ppl) - solo attendees meet people instead of eating alone - simpler org and cheaper tickets - eat better food (and local food when traveling)

I’m afraid this is just getting started. Supply chain hygiene is needed more than ever.

I have a Laravel app running serverless with Bref and generating PDF via chrome on Lambda. I added X-Ray to visualize it: 1 - the HTTP route (Laravel controller) invokes the pdf function synchronously and passes it a URL to render

I just realized that Laravel doesn't recognize "prod" (env name) as _production_ New Bref docs to address it: bref.sh/docs/larave...

I'm rewriting externals.io to Laravel, it's long overdue: github.com/mnapoli/ext...

L'IA a débarqué dans le quotidien des devs, il est dont évident que le Forum PHP 2026 en aura au programme. Mais quels sont les axes qui vous intéressent le plus autour du sujet ? Partagez avec nous vos attentes, vos envies, vos idées tout comme vos craintes, dans ce sondage anonyme.

I may have picked the wrong time to try Cursor again but I'm very pleasantly surprised! - very good app - picked up my Claude Code skills/commands - one subscription, all models - local projects, worktrees, cloud agents, all in one

New UI in Bref Cloud to go through Laravel failed jobs: list, retry, clear… bref.sh/cloud

I've been deep into genealogy research for the past year. As I collect more info, I want to render each family in my tree as a nice diagram. I built a free website that builds family trees and exports as images: family-tree-maker.mnapoli.fr It also works with Claude/ChatGPT via MCP

Updated to my /address-pr-review skill: - fixes are automatically pushed to the PR - threads are marked as resolved (unless Claude disagrees: it will explain why) Btw this skill is awesome to let Claude review and other Claude fix, highly recommend github.com/mnapoli/ski...

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?