Jetstreampro Laravel Package

Getting Started

Jetstream Pro is a commercial add-on that layers extra features onto a stock Laravel Jetstream application. v3.5.13 is a maintenance release — no new features, no API changes — so it's a safe drop-in if you're already on the 3.5 line.

1. composer require jetstreamlabs/jetstreampro on a Laravel 9.x app that already has Jetstream installed.
2. Note the new dependency floor: laravel/jetstream ^2.16.0 and laravel/framework ^9.50.2. If you were pinned below those, bump them first — Composer will otherwise resolve an older Jetstream Pro.
3. Publish the package's config and assets (php artisan vendor:publish and pick the jetstreampro-* tags — inspect the service provider's $publishes array for the exact tags in your version). Configure features in config/jetstreampro.php.
4. Run migrations, then npm install && npm run dev. The frontend is Vue 3 + Inertia compiled through Vite (laravel-vite-plugin) with Tailwind 3.3.

First use case: drop it into an existing Jetstream app, enable one pro feature in config, migrate, and confirm the corresponding Inertia page renders — that single vertical slice validates config publishing, asset building, and routing at once.

Implementation Patterns

• Layer, don't fork. Keep your Jetstream customizations in app/Actions/Jetstream and your own resources/views. Let the package own the pro surface. Override by publishing views/components rather than editing vendor/.
• Feature-flag via config. Enable pro features per-environment in config/jetstreampro.php so local/staging can run a reduced surface while prod gets the full set. It also makes it trivial to bisect a regression to one feature.
• Frontend overrides. The package ships Vue 3 SFCs; publish and edit the copies in resources/js/Pages / resources/js/Components to customize behavior. Keep Tailwind content globs pointed at wherever those land.
• Testing. This release moves the package's own suite to PHPUnit 10 (from 9.5). Your app tests are unaffected, but if you contribute upstream or copy their test scaffolding, write PHPUnit 10-style tests (#[Test] attributes, no --verbose-dependent docblock annotations).
• Contribution / CI. Upstream switched from StyleCI CLI to Pint. If you send a PR, run ./vendor/bin/pint first so the diff isn't whitespace noise.
• Upgrade cadence. Because releases like this are pure dependency churn, pin the package with ~3.5.0 and let patch bumps flow; review minor bumps (3.6+) manually.

Gotchas and Tips

• Compiled output can shift without any source change. This release bumps tailwindcss 3.2.4 → 3.3.2, axios 1.2.6, and webpack 5.82. Tailwind 3.3 changed some utility/color output — always npm run build and visually diff key pages after upgrading, even for a "maintenance" release.
• Published assets do not auto-update. The single most common "the fix didn't apply" cause is a stale published view/component overriding the vendor copy. Diff your published files against vendor/jetstreamlabs/jetstreampro/ after every upgrade.
• Tailwind content globs. If pro components render unstyled, your tailwind.config.js content array is almost certainly missing the vendor path. Safelist any dynamically composed classes.
• Vue version skew. The package pins [email protected] / @vue/[email protected]. Your app's Vue must be the same minor or you'll get Inertia hydration mismatches that look like unrelated JS bugs.
• Dev-dependency majors are not your problem. phpunit/phpunit 10.x, unplugin-auto-import 0.15, unplugin-icons 0.16, and the styleci/cli removal are all package-internal. Don't copy their package.json/composer.json dev blocks into your app.
• Cache clears are mandatory. After any upgrade: php artisan optimize:clear then re-cache. Config and compiled views hold stale class references.
• Debugging pro features. When a feature silently no-ops, check in this order: config flag enabled → provider registered → migrations run → published asset isn't shadowing vendor. That ordering resolves ~90% of support-style issues.
• Extension points. The config file and published views are the supported seams. Hooking the service provider's container bindings works but is unsupported and breaks on minor bumps.