Skip to main content
The Conversation API allows external access to retrieve, delete, and tag conversations within Rasa Studio.

Requirements

Quick Reference


Queries

rawConversations

Retrieves conversations for an assistant with optional filtering by time range and tags. Schema:
Input: RawConversationsInput Output: RawConversationsOutput Example:

conversationTags

Retrieves all conversation tags available for a given assistant. Schema:
Input: ConversationTagsInput Output: [ConversationTag!]! Returns a list of conversation tags: Example:

Mutations

addConversationTagToConversation

Adds a tag to a conversation. Schema:
Input: AddConversationTagToConversationInput Output: Returns true if the tag was successfully added.
If the tag already exists on the conversation, the endpoint will also return true. If a non-existent conversation, tag, or resource mismatch occurs, a descriptive error message with an appropriate error code is returned.
Example:

removeConversationTagFromConversation

Removes a tag from a conversation. Schema:
Input: RemoveConversationTagFromConversationInput Output: Returns true if the tag was successfully removed.
If the provided tagId is not linked to the conversationId, or if the provided conversationId or tagId is not present in the system, the endpoint will still return true.
Example:

deleteConversations

Deletes one or more conversations. Schema:
Input: DeleteConversationsInput Output: Returns the count of deleted conversations.
The count only includes conversations that were actually deleted. If all IDs passed to the API were invalid, count will be zero.
Example:

Reference

Conversation Events Structure

The conversationEvents field in rawConversations contains a list of events that occur within a conversation. These events follow a specific structure that varies by event type. For a full description of all possible events, see the Rasa Events Documentation.

User Uttered Event

Bot Uttered Event

Action Executed Event