PHP Script to augment the page content through AWS Bedrock with Claude Sonnet

wp-evereven_augment.php

<?php

// Ensure this script its used by consola or exit process
if (php_sapi_name() !== 'cli') {
    die('This script can only be run from the command line.');
}

// Database config (get all this data from wp-config.php)
$db_host = '<db_ip>:3306';
$db_name = '<db_name>';
$db_user = '<db_user>';
$db_pass = '<db_password>';
$table_prefix = 'wp_';

// AWS config (make sure to follow least privilege principle)
$aws_key = '********************';
$aws_secret = '****************************************';
$aws_region = 'us-east-1';

// Connect to datatase
$conn = new mysqli($db_host, $db_user, $db_pass, $db_name);
if ($conn->connect_error) {
    die("Connection failed: " . $conn->connect_error);
}

// Load SDK de AWS
require 'vendor/autoload.php';
use Aws\Exception\AwsException;
use Aws\BedrockRuntime\BedrockRuntimeClient;

// Setup Bedrock client
$bedrockClient = new BedrockRuntimeClient([
    'version' => 'latest',
    'region' => $aws_region,
    'credentials' => [
        'key' => $aws_key,
        'secret' => $aws_secret,
    ],
]);

// Optional function to filter the pages to interact
function isPageWithSomeCriteria($content, $language) {
    // custom implementation
    return true;
}

// Función para generar contenido con Claude Sonnet
function generateContent($prompt) {  
  global $bedrockClient;
  $messages = [
    [
      'role' => 'user',
     'content' => $prompt
    ]
  ];
  
  try {
    $result = $bedrockClient->invokeModel([
      'modelId' => 'anthropic.claude-3-sonnet-20240229-v1:0',
      'contentType' => 'application/json',
      'accept' => 'application/json',
      'body' => json_encode([
        'anthropic_version' => 'bedrock-2023-05-31',
        'max_tokens' => 1000,
        'messages' => $messages
      ])
    ]);
    
    $response = json_decode($result->get('body')->getContents(), true);
    return $response['content'][0]['text'];
  } catch (AwsException $e) {
    echo "Error al invocar el modelo de Bedrock: " . $e->getMessage() . "\n";
    return false;
  }
}

// SQL query to get all published pages in worpdress (omitting drafts and hidden)
$sql = "SELECT p.ID, p.post_title, p.post_content, p.post_name, p.post_parent,
               (SELECT meta_value FROM {$table_prefix}postmeta WHERE post_id = p.ID AND meta_key = '_wp_page_template') as template,
               (SELECT t.slug FROM {$table_prefix}terms t
                JOIN {$table_prefix}term_taxonomy tt ON t.term_id = tt.term_id
                JOIN {$table_prefix}term_relationships tr ON tt.term_taxonomy_id = tr.term_taxonomy_id
                WHERE tr.object_id = p.ID AND tt.taxonomy = 'language') as language
        FROM {$table_prefix}posts p
        WHERE p.post_type = 'page' AND p.post_status = 'publish'
        ORDER BY p.ID";

$result = $conn->query($sql);
$processed_pages = 0;

// Iterate over the pages
if ($result->num_rows > 0) {
    while($row = $result->fetch_assoc()) {
        echo "id=" . $row["ID"] . " -----------------------------------------------\n";
        echo "link: /" . $row["post_name"] . "\n";
        echo "title: " . $row["post_title"] . "\n";
        echo "parent: " . ($row["post_parent"] != 0 ? $row["post_parent"] : "none") . "\n";
        echo "language: " . ($row["language"] ? $row["language"] : "N/A") . "\n\n";

        if (isPageWithSomeCriteria($row["post_content"], $row["language"])) {
            
          $prompt = "Extend the content of this page in a fancy way that bla bla bla....:\n\n" .
                    "Title: " . $row["post_title"] . "\n" .
                    "Content: " . strip_tags($row["post_content"]) . "\n\n" .
                    "The structure must be ...\n" .
                    "Do not forget to include ...\n" .
                    "The output language must be: " . ($row["language"] == 'es' ? "español" : "inglés") . ".\n";

          $response = generateContent($prompt);
          
          // Outputs model response
          if ($response) {
            echo $response . "\n\n";
            $processed_pages++;
          
          } else {
              echo "Unable to generate content for this page.\n\n";
          }

          // Manual Throttling
          sleep(5);
        } else {
          echo "The page doesn't follow criteria, so it's discarded\n\n";
        }
    }
} else {
    echo "No pages found.";
}

// Close connection and report
$conn->close();
echo "Total number of pages processed: " . $processed_pages . "\n";

?>