larswiegers / laravel-ai-evaluation

0.1.6 2026-09-13 13:21 UTC

Requires

• php: ^8.3
• illuminate/console: ^12.0|^13.0
• illuminate/support: ^12.0|^13.0
• laravel/ai: >=0.5.0 <0.12.0

Requires (Dev)

• orchestra/testbench: ^11.0
• pestphp/pest: ^4.5
• phpunit/phpunit: ^12.0

MIT

• Lars Wiegers

• testing
• laravel
• ai
• llm
• evals

Published on Packagist.org on 2026-09-13 13:22 UTC

README

Test real Laravel AI agent behavior with repeatable evals that run through Artisan and CI.

Quick start

composer require --dev larswiegers/laravel-ai-evaluation
php artisan make:ai-evals refund-policy
php artisan ai-evals:run

use LaravelAIEvaluation\AIEval;

it('answers refund policy questions', function () {
    AIEval::agent(App\Ai\Agents\SupportAgent::class)
        ->input('Can I get a refund?')
        ->expectContains(['refund', '30 days'])
        ->run()
        ->assertPasses();
});

For installation, expectations, standalone reports, and CI output formats, see the docs:

• Documentation