In `@pkg/subscraping/sources/urlscan/urlscan.go`:
- Around line 60-66: The Run method resets s.errors, s.results, and s.requests
but not s.skipped, causing stale stats; update the Run(ctx context.Context,
domain string, session *subscraping.Session) method to set s.skipped = false at
the start alongside the other counters so Statistics() reflects the current run
state (modify the Source.Run function where s.errors/s.results/s.requests are
initialized).
- Around line 210-244: The current request loop returns resp,nil for non-OK
statuses (when err==nil), causing callers like enumerate to decode error
payloads as success; update the logic in the retry block (the code handling
resp, attempt, maxRetries, backoff in urlscan.go) so that for any resp with
StatusCode != http.StatusOK and not a retryable status (not 429/503) you call
session.DiscardHTTPResponse(resp) and return a non-nil error (e.g.
fmt.Errorf("unexpected status %d", resp.StatusCode)) instead of returning
resp,nil; keep existing retry behavior for 429/503 using the
X-Rate-Limit-Reset-After header, ctx cancellation, and exponential backoff.

In `@pkg/subscraping/sources/urlscan/urlscan.go`:
- Around line 210-259: The final "return nil, fmt.Errorf(\"max retries
exceeded\")" after the retry loop is unreachable because every path inside the
for attempt := 0; attempt <= maxRetries; attempt++ loop returns; remove that
trailing return to clean up dead code (delete the return nil, fmt.Errorf("max
retries exceeded") line in pkg/subscraping/sources/urlscan/urlscan.go) or
alternatively change the loop to a strict < maxRetries form if you intend a
post-loop fallback; target the retry loop using the session.Get / attempt logic
and the trailing return statement for the fix.