Code-Cube's platform covers multiple features or modules:
- GTM client/server-side monitoring, cloud run container monitoring, anomaly detection
- DataLayer Guard (front-end object monitoring), scenario monitoring
- Network Inspector
What’s the difference between DataLayer Guard and Network Inspector?
Data Layer Guard checks the full DOM and any front-end JSON objects, validating event schemas, field requirements, data types, and URL configurations; JSON import is supported for convenience.
- Scenario Builder can simulate full user journeys on web (e.g., login, cookie acceptance, clicking fields) using a WebKit-based browser engine; native app scenarios are not supported in the same way.
Network Inspector enables deep network-level inspection of HTTP-, XHR requests and WebSockets, including custom headers and endpoints. The purpose is to monitor analytics data and HTTP requests in the network layer distinct from the DOM-level Data Layer Guard.
Technical Overview: Data Layer vs. Network Layer Monitoring
1. Architectural Distinction The primary difference between these two systems lies in the Data Source. While both serve the purpose of validating analytics and application health, they intercept data at different points in the browser's architecture.
2. Validation & health monitoring
While the user interface remains consistent across both modules to ensure a seamless user experience, the Assertion Logic differs:
- Data Layer Assertions: Focuses on the presence of specific keys and values within the JSON objects pushed to the data layer (e.g., checking if
event: "page_view"exists). - Network Assertions: Focuses on the "health" of the communication. This includes:
- Status Code Verification: Ensuring the request returns a
200 OKor other expected status. - Header Validation: Confirming that mandatory HTTP headers are present and contain correct metadata.
- Payload Inspection: Verifying the data being sent directly to the analytics server before it is processed.
3. Use case & synergy
As noted in the discussion, these tools are generally treated as mutually exclusive based on the client's implementation:
- Standard Implementation: Most clients use DOM-based tracking (DataLayer).
- Unique/Legacy Implementation: Specific clients rely on direct network-level tracking, necessitating the Network Litigator's specialized code to capture metadata that never reaches the DOM.