I'm facing a puzzling issue with Webex bots integrated with Jira, deployed on-prem behind an F5 load balancer. Our setup routes public HTTPS endpoints (e.g., https://<our-company-domain>/botA to backend port xxxx1, and /botB to port xxxx2) via F5 to internal Flask servers. I have two different bots: an older Jira Bug Bot (developed about five months ago) and a new SID bot, both using adaptive cards and similar code and configurations. The Jira Bug Bot works only on port xxxx1 (/botA), but fails completely when moved to port xxxx2 (/botB), even with identical code, config, and verified routing. The new SID bot, meanwhile, does not fully work on either port: it can respond to plain text messages (e.g., replying to "Hi" with the filter card), but adaptive card actions (attachmentActions such as "Search" button clicks) do not reach the Flask server—there is no evidence of a POST, either in my Python logs or on F5 traffic monitors. Regular text message POSTs show up normally on both ports, confirming F5 is routing those as expected. I have explicitly registered Webex webhooks with the correct URLs, deleted and recreated them multiple times, and even tried different libraries (webexteamsbot and webexteamssdk), but the issue remains. To summarize: adaptive card actions only work on port xxxx1 for the old bot, never on xxxx2, and not at all for the new bot on either port, despite the same backend logic, configuration, and routing for all. Has anyone else encountered selective POST failures for attachment actions in this kind of setup, and what troubleshooting or infrastructure changes would you recommend?
I am manly using the webexteamsbot python library
