Front End & Back End Automation

Context

During automation testing, a script intended to validate end-to-end data flow was interrupted at the boundary between frontend UI interactions and backend API processing.

Observed Behavior

• The automation script successfully completed front-end interaction steps (form fill, button trigger).
• The script halted immediately after submitting the request — no response was captured, and no backend log entry was created.
• Manual execution of the same flow produced the expected result without error.

Investigation Steps

1. Compared the automation script's HTTP headers with those of the manual request — identified a missing Content-Type header in the automated POST body.
2. Reviewed the backend endpoint's request validation logic — it was rejecting requests without the explicit application/json content type silently (returning a 400 with no body).
3. Inspected the CI pipeline logs to confirm whether the issue was environment-specific — it was reproducible only in the headless browser environment.

Root Cause

The headless browser automation framework did not automatically inject Content-Type: application/json when making fetch-based API calls, unlike the standard browser which inferred it from the request body format.

Resolution

• Added an explicit header configuration step to the automation script before any API-triggering UI interactions.
• Proposed adding a middleware logging layer to the backend to surface rejected requests with descriptive error messages instead of silent 400s.
• Documented the fix in the knowledge base under "Automation Gotchas".

Lesson Learned

Automation environments can subtly differ from real browser environments in ways that cause silent failures. Always validate automation-driven requests against the raw HTTP layer, not just the UI outcome.