Solving bot issues
Whenever your bot does not behave the way you expect it to, you will need to investigate to see exactly what the issue is. This is also called debugging. But where to start with debugging? How can you find what the cause of an issue is? This article will go through the basics of debugging in the Chatlayer platform. Make sure to check the subpages on the left side, where some common issues in the Chatlayer platform are described and how to solve them.
Debugging can be done very easily in the emulator, this is the
Test your bot
window in the right bottom corner. The emulator consists of three parts:The main part of the emulator shows the bot response on the left side, and typed input from the user on the right side:

The following symbols are available for debugging:

Open the specific bot dialog that gives that rep

Repeat that particular bot dialog or user input

Add that user input as an expression
Next to that, you can find the following options in this window:

Open the Debugger tab

Open the NLP Result tab

Clear the entire conversation history and saved variables and start from scratch

Choose the type of debugger window. Recommended is Debugger for the most information

Choose the language for this session in the emulator
In the debugger you can find the following information:

User says
the input the user has givenIntent
the recognized intentBot dialog
with the id and name of the current dialogRead-Only Session data
This is all data that is retrieved up until this point in the session. If you need to use any of these variables, it is important that you add internal.
before the variable. For example internal.channel
If there are any variables collected in the conversation along the way, they will be listed at the bottom. If you want to use these,
internal.
is not necessary
If you need to use any of the variables with a
{ }
behind it, remember to use the correct variable name when you open the variable. For example internal.currentDialogstate.name
Below
Messages
(on the right side) you can see all the messages the bot has given up until that point.The NLP Result tab shows the result of the NLP engine. Any recognized intents or entities are shown here

This expression gives a 93.86% positive score for the Book train ticket intent
This information helps locating intent issues - if a wrong response is given in the bot, you can check here which intent was triggered.

The word 'fries' gives a 100% match with the value 'fries' of the 'food' entity
Last modified 2yr ago