Feature Request

Is your feature request related to a problem? Please describe.
No

Describe the suggestion or request in detail
The following issues (mostly speech-related) should have more tests written:

• Should not stop after first spoken text #2016
• Fix mobile focus UX #2087
• Remove markdown from spoken text for screen reader #2096
  • Not trivial to test aria-label
• Fix Speech speaking the title, subtitle, and text of cards #2118

Also, the mockWebSpeech.js file needs more coverage in these areas:

• Web Chat should select voice based on language
• Web Chat should set SpeechRecognition.lang based on language
• Unhappy paths
  • Bird tweets
  • User denied access to microphone
  • Unrecognizable voice (a.k.a. no-match)
  • Network error or wrong credentials
• Abort while recognizing, then restart microphone again (regression of Speech: Stop microphone button should re-enable the button after a few seconds #2240)
• TBD

Speech tests around consecutive activities with different inputHint

Note that 1st and 2nd activity could sometimes have a slight delay between them, ~1 second.

Reliability issues

• suggested-actions command should show suggested actions with images
  • After emptycard was sent, we didn't wait until suggested actions show up
  • Today, we use minNumActivitiesShown to test, but this cannot be used to test empty card or suggested actions
  • Use suggestedActionsShowed condition to wait until suggested actions show up before taking the snapshot

Describe alternatives you have considered
N/A

[Enhancement]