WhatsApp Business Bot - Bot Details API | Developer Documentation

Start with the reply that was actually sent

Use this diagnostic when a customer receives an unexpected or repeated reply through a connected tool or your application. A Meta command or ice breaker is an input component; the CRM/help desk or custom service may have separate rules that act on incoming messages. The imported Bot Details reference does not establish a ChatArchitect automation-management endpoint.

Record one example before changing rules

• The business number, customer conversation and incoming message or action preceding the reply.
• The exact unexpected reply, time with time zone and available outgoing attempt/result reference.
• The connected tool/workspace and any known rule, job or custom application that could have sent it.
• Whether there are two separate outgoing attempts or one item shown more than once.

Use the attempt timeline and results to distinguish those observations. An extra callback or duplicated UI item does not by itself prove that another WhatsApp message was sent.

Trace the sending source

1. Inspect the available sending record in the connected tool or application for that particular attempt.
2. If it shows a rule or scheduled job, ask its owner to compare the trigger, recipient and connection with the observed message.
3. For your own service, compare its incoming handler, processing outcome and recorded outbound call using the incoming-text guide and the Quick Start.
4. Check whether another authorized workflow uses the same business number. Use the routing inventory rather than guessing the source from a similar display name.

For a component that produces no useful response, use prompt troubleshooting; configuring commands or ice breakers in WhatsApp Manager does not establish a reply rule in the CRM. Check the supported actions of the actual tool instead of treating a Meta bot identifier as a ChatArchitect rule identifier.

Correct the identified workflow and retest

Use the owner's supported controls to correct or pause the identified rule. Keep the incoming route and other established workflows intact unless the agreed change requires otherwise. Then repeat a permitted test and compare the number of actual outgoing attempts and the replies received by the contact.

Do not assume automatic-reply controls exist in every integration or that all platforms use the same retry/duplicate rules. If the source remains unclear, send the labelled example and configuration references to support with secrets and unrelated private content removed.