Hi Facu,
Apologize for the delay. I've been dedicated full-time to priority tasks since Tuesday last week.
I don't see anything that should be causing these delays, but bear in mind the alert is actually an agent queue flood, and as such, it is indicating that the agent is generating more events than it can send to the Wazuh Manager.
Integrations running on the Manager are run in a queue, so if the integration is getting triggered by a very large number of events that could be locking new reports from being sent.
It is also possible that there is a connectivity issue taking several connection attempts to actually reach the telgram/e-mail servers.
It would probably be a good idea to add logging/debugging facilities to your scripts and check their output, just to check the execution timestamp (when the Wazuh Server actually runs the scripts) and to gather any useful data on what is causing the delays.
Let me know if you've gathered any further insight in this topic.
Regards,
Federico