I’m getting multiple duplicates of messages on Slack (same actions gets triggered).
When my custom action gets called I get
200 from ngrok, but right after that I get
500 and a
http_timeout error. Also I’ve added
errors_ignore_retry: | - "http_timeout"
slack in my
credentials.yml which doesn’t seem to work.
Bot is working properly on Rasa shell.
http_timeout I also get this error:
 [ERROR] Exception occurred while handling uri: 'http://xxxxxxxxxxxx.ngrok.io/webhooks/slack/webhook' Traceback (most recent call last): File "e:\anaconda\envs\chatbot_env\lib\site-packages\sanic\app.py", line 931, in handle_request response = await response File "e:\anaconda\envs\chatbot_env\lib\site-packages\rasa\core\channels\slack.py", line 526, in webhook return await self.process_message( File "e:\anaconda\envs\chatbot_env\lib\site-packages\rasa\core\channels\slack.py", line 351, in process_message return response.text( File "e:\anaconda\envs\chatbot_env\lib\site-packages\sanic\response.py", line 224, in text raise TypeError( TypeError: Bad body type. Expected str, got NoneType)
And a side questions: does
utter_message(text) have a string length limit?
Could the errors that am getting be in any way related to new line tokens
\n or emojis inside responses?
Also for some reason, half of the response text doesn’t get printed on Slack.
I have few
JSON files that contain answers to questions asked by users. In custom actions I open these files and go through them in a for loop (nothing fancy) and concatenate them in one big
message variable which is then given to the
text= argument of
utter_message method. Besides that, I followed Rasa guide for deploying bot to Slack so everything is the same except
errors_ignore_retry that I mentioned earlier.