Fires a burst of concurrent requests to an external origin (mimicking a chatty PMS page) alongside a single named target call, to test whether tshark's capture/decode pipeline reliably sees every connection under concurrent HTTP/2 load. Must be served over TLS/HTTP2 to actually exercise the decode path being tested. The burst target must not be *.cloudcruise.com/.app — that's hard-skipped by design (see the control section below).
(not fired yet)
(not fired yet)
This control hits demo.cloudcruise.com's own API, which matches cc_tshark_event_mapper.py's hard-skip-host regex for *.cloudcruise.com. It exists to prove that filter is working — if an EXTRACT_NETWORK node ever captures this (0 control noise requests fired alongside it), that's a regression in the hard-skip list, not evidence about the decode race this page is otherwise testing.
Success on this page (fetches resolve without throwing) only proves the browser sent and received everything — it says nothing about capture. Compare against what an EXTRACT_NETWORK workflow node targeting the marker URL actually sees, and against the worker's tshark /stats endpoint (tls_handshakes_in_keylog vs tls_connections_decoded) to see whether the decode gap reproduces.