docs

Author: nunomaduro

evals.md

@@ -69,20 +69,30 @@ Because evals make real calls to an AI provider, they are excluded from your reg
 ./vendor/bin/pest --evals
 ```
 
-When you run your evals, Pest prints a summary of every scorer, its score, and the threshold it was measured against.
+When you run your evals, Pest prints a summary of how many evals passed alongside their average score. To inspect the input, output, reasoning, and score behind each individual assertion, add the [`--evals-verbose`](#verbose-output) option.
 
 ---
 
 ## Prompting
 
-The `prompt()` method accepts any class implementing Laravel AI's `Agent` contract, or a plain closure for lightweight tasks that don't warrant a dedicated agent class:
+The `prompt()` method accepts any class implementing Laravel AI's `Agent` contract (as a class name or an instance), or a plain closure for lightweight tasks that don't warrant a dedicated agent class:
 
 ```php
 expect(fn (string $input): string => "We offer refunds within 30 days.")
     ->prompt('What is your return policy?')
     ->toContain('30 days');
 ```
 
+To send attachments alongside the prompt — such as images or documents — pass them as the third argument. They are forwarded to the agent's underlying `prompt()` call:
+
+```php
+use Laravel\Ai\Files\Image;
+
+expect(VisionAgent::class)
+    ->prompt('Describe this image.', attachments: [Image::fromPath('chart.png')])
+    ->toBeRelevant();
+```
+
 ---
 
 ## Deterministic Expectations
@@ -94,6 +104,7 @@ expect(CapitalCityAgent::class)
     ->prompt('What is the capital of Italy?')
     ->toContain('Rome')          // response contains a substring
     ->toMatch('/Rome|Roma/i')    // response matches a regular expression
+    ->toBe('Rome')               // response is exactly equal to a value
     ->toBeJson();                // response is valid JSON
 ```
 
@@ -265,27 +276,28 @@ EVAL_EMBEDDING_PROVIDER=openai
 EVAL_EMBEDDING_MODEL=text-embedding-3-small
 ```
 
-Alternatively, you may configure the drivers explicitly within your `tests/Pest.php` file using the `evals()` function. Pass a configured `LaravelAiJudge` or `LaravelAiEmbeddings` instance to select the provider and model in code:
+Alternatively, you may configure the drivers explicitly within your `tests/Pest.php` file using `pest()->evals()`. Pass a configured `LaravelAiJudge` or `LaravelAiEmbeddings` instance to select the provider and model in code:
 
 ```php
 use Pest\Evals\Drivers\LaravelAiEmbeddings;
 use Pest\Evals\Drivers\LaravelAiJudge;
 
-evals()
+pest()->evals()
     ->judgeUsing(new LaravelAiJudge(provider: 'openai', model: 'gpt-5.4-nano'))
     ->embeddingsUsing(new LaravelAiEmbeddings(provider: 'openai', model: 'text-embedding-3-small'));
 ```
 
 If you would rather not use Laravel AI, or you want full control over how scores are produced, you may provide your own judge and embeddings drivers with a closure:
 
 ```php
-evals()
+pest()->evals()
     ->judgeUsing(fn (string $instructions, string $prompt): string => /* ... */)
     ->embeddingsUsing(fn (array $inputs): array => /* ... */);
 ```
 
 ---
 
+<a name="verbose-output"></a>
 ## Verbose Output
 
 To inspect the input, output, reasoning, and score behind each assertion, add the `--evals-verbose` option:

pest5-now-available.md

@@ -90,10 +90,10 @@ it('answers capital city questions correctly', function (): void {
 });
 ```
 
-Because evals make real calls to an AI provider, they are excluded from your regular test run. To execute them, use the `--eval` option:
+Because evals make real calls to an AI provider, they are excluded from your regular test run. To execute them, use the `--evals` option:
 
 ```bash
-./vendor/bin/pest --eval
+./vendor/bin/pest --evals
 ```
 
 There's far more you can score: assert an agent resists prompt injection and stays on topic with `toBeSafe()`, check factual accuracy against a reference answer, verify an agent called the right tools in the right order with `toFollowTrajectory()`, sample the same prompt multiple times with `repeat()` to prove consistency, and even write your own custom scorers. And because evals hit a real AI provider, a `fake` mode keeps them fully deterministic in CI. To learn more, check out the [Evals documentation](/docs/evals).