如何处理对话流中自定义有效负载的答案？

Question

I have a bot that handles well the welcome intent using Dialogflow for Slack. However I don't know how to deal with the answer to the welcome intent to fire the second post. Indeed, the welcome intent , which outputs a await_answer1 context, shows the following template in json:

{
  "slack": {
    "text": "",
    "attachments": [
      {
        "blocks": [
          {
            "type": "section",
            "text": {
              "type": "mrkdwn",
              "text": "*How have you been feeling?* Poll by <fakeLink.toUser.com|Mihailo>"
            }
          },
          {
            "type": "divider"
          },
          {
            "type": "section",
            "text": {
              "type": "mrkdwn",
              "text": ":tired_face: *I’ve been feeling more exasperated and hopeless*"
            },
            "accessory": {
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              },
              "value": "1"
            }
          },
          {
            "accessory": {
              "value": "2",
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              }
            },
            "type": "section",
            "text": {
              "type": "mrkdwn",
              "text": ":expressionless: *Generally, less freaked out than other people*"
            }
          },
          {
            "accessory": {
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              },
              "value": "3"
            },
            "type": "section",
            "text": {
              "type": "mrkdwn",
              "text": ":relieved: *More calm and hopeful*"
            }
          },
          {
            "type": "divider"
          },
        ]
      }
    ]
  }
}

And I would like to handle the answer. So I created an answer1 intent which takes the await_answer1 as an input context. The training phrases are the output of the above template: 1 , 2 , 3 , 4 , 5 . And the Default text response is Interesting! However after selecting the answer is the Fallback intent rather than answer1 . Therefore, how to handle thee answer of a custom payload in dialogflow?

AngelDev answer

I tried to include block_id :

{
  "slack": {
    "text": "",
    "attachments": [
      {
        "blocks": [
          {
            "type": "section",
            "text": {
              "type": "mrkdwn",
              "text": "*How have you been feeling?* Poll by <fakeLink.toUser.com|Mihailo>"
            }
          },
          {
            "type": "divider"
          },
          {
            "accessory": {
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              },
              "value": "1"
            },
            "type": "section",
            "block_id": "1",
            "text": {
              "type": "mrkdwn",
              "text": ":tired_face: *I’ve been feeling more exasperated and hopeless*"
            }
          },
          {
            "accessory": {
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              },
              "value": "2"
            },
            "type": "section",
            "block_id": "2",
            "text": {
              "type": "mrkdwn",
              "text": ":expressionless: *Generally, less freaked out than other people*"
            }
          },
          {
            "type": "section",
            "text": {
              "type": "mrkdwn",
              "text": ":relieved: *More calm and hopeful*"
            },
            "accessory": {
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              },
              "value": "3"
            }
          },
          {
            "type": "section",
            "block_id": "4",
            "text": {
              "type": "mrkdwn",
              "text": ":fearful: *More scared and panicked*"
            },
            "accessory": {
              "value": "4",
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              }
            }
          },
          {
            "type": "section",
            "block_id": "5", 
            "text": {
              "type": "mrkdwn",
              "text": ":open_mouth: *More surprised and baffled*"
            },
            "accessory": {
              "value": "5",
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              }
            }
          }
        ]
      }
    ]
  }
}

Yet when I click on the button the DialogFlow bot falls on the DefaultFallback intent rather than going to the answer1 intent that should handle the await_answer1 context.

Answer 1

I'm not completly sure why, but if you remove the output and input context, this setup works fine.

These are the accessory properties I used, I only changed the value property and added those values as a example phrase to my intents.

"accessory": {
              "value": "First",
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              }
            },

"accessory": {
              "value": "Second",
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              }
            },

"accessory": {
              "type": "button",
              "text": {
                "emoji": true,
                "type": "plain_text",
                "text": "Vote"
              },
              "value": "Third"

If you are looking to build a very customized Slack experience it might be worth to investigate into the Slack API . OnClick events and other Slack specific events are currenty not supported from within the Dialogflow UI, but you can integrate it with code in a webhook and from there customize your conversation in greater detail.

Answer 2

You must include a block_id in your custom payload in each element of the block arrangement and this value must be in the training phrases of your intention since when you receive an interaction you receive this block_id. For example

  {
      "type": "section",
      "text": {
          "type": "mrkdwn",
          "text": "*How have you been feeling?* Poll by <fakeLink.toUser.com|Mihailo>"
        },
       block_id: "section_1"
     }

I hope it is what you are looking for but you can write to me to help you solve your doubts.