crewli/api/app/Services/Observability/SentryEventScrubber.php

<?php

declare(strict_types=1);

namespace App\Services\Observability;

use Sentry\Event;
use Sentry\EventHint;
use Symfony\Component\HttpKernel\Exception\HttpException;

/**
 * PII scrubber registered as Sentry's `before_send` hook (RFC-WS-7 §3.7).
 *
 * Two responsibilities:
 *  - Drop sub-500 HttpExceptions: ignore_exceptions in config/sentry.php is
 *    class-only; status-based filtering must happen here.
 *  - Strip sensitive request data from outgoing events: body keys, headers,
 *    query string parameters, and form_values payloads (definitionally PII
 *    in Crewli — entire payload is replaced wholesale).
 */
final class SentryEventScrubber
{
    private const SENSITIVE_BODY_KEYS = [
        'password', 'password_confirmation', 'current_password',
        'token', 'api_key', 'secret', 'webhook_secret', 'dsn',
        'signature', 'authorization', 'cookie', 'bearer',
        'iban', 'bic', 'passport_number', 'bsn',
    ];

    private const SENSITIVE_HEADERS = [
        'authorization', 'cookie', 'set-cookie',
        'x-api-key', 'x-impersonation-token',
    ];

    private const SENSITIVE_QUERY_KEYS = [
        'token', 'api_key',
    ];

    private const SCRUBBED = '[scrubbed]';

    private const FORM_VALUES_KEY = 'form_values';

    private const FORM_VALUES_REPLACEMENT = '[scrubbed_form_values]';

    private const MAX_DEPTH = 10;

    public function scrub(Event $event, ?EventHint $hint = null): ?Event
    {
        if ($hint?->exception instanceof HttpException && $hint->exception->getStatusCode() < 500) {
            return null;
        }

        $request = $event->getRequest();

        if ($request !== []) {
            $event->setRequest(array_merge($request, [
                'data' => $this->scrubBody($request['data'] ?? []),
                'headers' => $this->scrubHeaders($request['headers'] ?? []),
                'query_string' => $this->scrubQueryString($request['query_string'] ?? ''),
                'cookies' => self::SCRUBBED,
            ]));
        }

        return $event;
    }

    /**
     * @param  mixed  $data
     * @return mixed
     */
    private function scrubBody($data, int $depth = 0)
    {
        if (! is_array($data)) {
            return $data;
        }

        if ($depth > self::MAX_DEPTH) {
            return ['[max_depth]'];
        }

        foreach ($data as $key => $value) {
            if (is_string($key) && strtolower($key) === self::FORM_VALUES_KEY) {
                $data[$key] = self::FORM_VALUES_REPLACEMENT;

                continue;
            }

            if (is_string($key) && in_array(strtolower($key), self::SENSITIVE_BODY_KEYS, true)) {
                $data[$key] = self::SCRUBBED;

                continue;
            }

            if (is_array($value)) {
                $data[$key] = $this->scrubBody($value, $depth + 1);
            }
        }

        return $data;
    }

    /**
     * @param  array<string, mixed>|string  $headers
     * @return array<string, mixed>|string
     */
    private function scrubHeaders($headers)
    {
        if (! is_array($headers)) {
            return $headers;
        }

        foreach (array_keys($headers) as $name) {
            if (in_array(strtolower((string) $name), self::SENSITIVE_HEADERS, true)) {
                $headers[$name] = self::SCRUBBED;
            }
        }

        return $headers;
    }

    private function scrubQueryString(string $queryString): string
    {
        if ($queryString === '') {
            return '';
        }

        parse_str($queryString, $parsed);

        foreach ($parsed as $key => $value) {
            if (is_string($key) && in_array(strtolower($key), self::SENSITIVE_QUERY_KEYS, true)) {
                $parsed[$key] = self::SCRUBBED;
            }
        }

        return http_build_query($parsed);
    }
}