- Resource: Example
- JSON representation
- PlaybookInput
- PlaybookOutput
- State
- Action
- UserUtterance
- Event
- AgentUtterance
- ToolUse
- DataStoreToolTrace
- WebhookToolTrace
- LlmCall
- RetrievedExample
- RetrievalStrategy
- TokenCount
- IntentMatch
- MatchedIntent
- FlowStateUpdate
- PageState
- FunctionCall
- PlaybookInvocation
- OutputState
- FlowInvocation
- PlaybookTransition
- FlowTransition
- TTS
- STT
- Span
- NamedMetric
- Status
- Exception
- Methods
Resource: Example
Example represents a sample execution of the playbook in the conversation.
An example consists of a list of ordered actions performed by end user or Dialogflow agent according the playbook instructions to fulfill the task.
JSON representation |
---|
{ "name": string, "playbookInput": { object ( |
Fields | |
---|---|
name |
The unique identifier of the playbook example. Format: |
playbookInput |
Optional. The input to the playbook in the example. |
playbookOutput |
Optional. The output of the playbook in the example. |
actions[] |
Required. The ordered list of actions performed by the end user and the Dialogflow agent. |
displayName |
Required. The display name of the example. |
description |
Optional. The high level concise description of the example. The max number of characters is 200. |
tokenCount |
Output only. Estimated number of tokes current example takes when sent to the LLM. |
createTime |
Output only. The timestamp of initial example creation. Uses RFC 3339, where generated output will always be Z-normalized and uses 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
updateTime |
Output only. Last time the example was updated. Uses RFC 3339, where generated output will always be Z-normalized and uses 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
conversationState |
Required. Example's output state. |
languageCode |
Optional. The language code of the example. If not specified, the agent's default language is used. Note: languages must be enabled in the agent before they can be used. Note: example's language code is not currently used in dialogflow agents. |
PlaybookInput
Input of the playbook.
JSON representation |
---|
{ "precedingConversationSummary": string, "actionParameters": { object } } |
Fields | |
---|---|
precedingConversationSummary |
Optional. Summary string of the preceding conversation for the child playbook invocation. |
actionParameters |
Optional. A list of input parameters for the action. |
PlaybookOutput
Output of the playbook.
JSON representation |
---|
{
"executionSummary": string,
"state": enum ( |
Fields | |
---|---|
executionSummary |
Optional. Summary string of the execution result of the child playbook. |
state |
End state of the playbook. |
actionParameters |
Optional. A Struct object of output parameters for the action. |
State
Playbook output state.
Enums | |
---|---|
STATE_UNSPECIFIED |
Unspecified state. |
OK |
Playbook succeeded. |
CANCELLED |
Playbook cancelled. |
FAILED |
Playbook failed. |
ESCALATED |
Playbook failed due to escalation. |
Action
Action performed by end user or Dialogflow agent in the conversation.
JSON representation |
---|
{ "displayName": string, "startTime": string, "completeTime": string, "subExecutionSteps": [ { object ( |
Fields | |
---|---|
displayName |
Output only. The display name of the action. |
startTime |
Output only. Timestamp of the start of the agent action. Uses RFC 3339, where generated output will always be Z-normalized and uses 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
completeTime |
Output only. Timestamp of the completion of the agent action. Uses RFC 3339, where generated output will always be Z-normalized and uses 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
subExecutionSteps[] |
Optional. The detailed tracing information for sub execution steps of the action. |
status |
Optional. Output only. The status of the action. |
Union field action . Action details. action can be only one of the following: |
|
userUtterance |
Optional. Agent obtained a message from the customer. |
event |
Optional. The agent received an event from the customer or a system event is emitted. |
agentUtterance |
Optional. Action performed by the agent as a message. |
toolUse |
Optional. Action performed on behalf of the agent by calling a plugin tool. |
llmCall |
Optional. Output only. LLM call performed by the agent. |
intentMatch |
Optional. Output only. Intent Match in flows. |
flowStateUpdate |
Optional. Output only. The state machine update in flows. |
playbookInvocation |
Optional. Action performed on behalf of the agent by invoking a child playbook. |
flowInvocation |
Optional. Action performed on behalf of the agent by invoking a CX flow. |
playbookTransition |
Optional. Action performed on behalf of the agent by transitioning to a target playbook. |
flowTransition |
Optional. Action performed on behalf of the agent by transitioning to a target CX flow. |
tts |
Optional. Text-to-speech action performed by the agent. |
stt |
Optional. Speech-to-text action performed by the agent. |
UserUtterance
UserUtterance represents one message sent by the customer.
JSON representation |
---|
{ "text": string, "audioTokens": [ integer ], "audio": string } |
Fields | |
---|---|
text |
Required. Message content in text. |
audioTokens[] |
Optional. Tokens of the audio input. |
audio |
Optional. Audio input. A base64-encoded string. |
Event
Event represents the event sent by the customer.
JSON representation |
---|
{ "event": string, // Union field |
Fields | |
---|---|
event |
Required. Name of the event. |
Union field payload . Payload of the event. payload can be only one of the following: |
|
text |
Optional. Unstructured text payload of the event. |
AgentUtterance
AgentUtterance represents one message sent by the agent.
JSON representation |
---|
{ "text": string, "requireGeneration": boolean } |
Fields | |
---|---|
text |
Required. Message content in text. |
requireGeneration |
Optional. True if the agent utterance needs to be generated by the LLM. Only used in webhook response to differentiate from empty text. Revisit whether we need this field or mark |
ToolUse
Stores metadata of the invocation of an action supported by a tool.
JSON representation |
---|
{ "tool": string, "displayName": string, "action": string, "inputActionParameters": { object }, "outputActionParameters": { object }, // Union field |
Fields | |
---|---|
tool |
Required. The |
displayName |
Output only. The display name of the tool. |
action |
Optional. Name of the action to be called during the tool use. |
inputActionParameters |
Optional. A list of input parameters for the action. |
outputActionParameters |
Optional. A list of output parameters generated by the action. |
Union field ToolTrace . The tracing information for different tools. ToolTrace can be only one of the following: |
|
dataStoreToolTrace |
Optional. Data store tool trace. |
webhookToolTrace |
Optional. Webhook tool trace. |
DataStoreToolTrace
The tracing information for the data store tool.
JSON representation |
---|
{
"dataStoreConnectionSignals": {
object ( |
Fields | |
---|---|
dataStoreConnectionSignals |
Optional. Data store connection feature output signals. |
WebhookToolTrace
The tracing information for the webhook tool.
JSON representation |
---|
{ "webhookTag": string, "webhookUri": string } |
Fields | |
---|---|
webhookTag |
Optional. The tag of the webhook. |
webhookUri |
Optional. The url of the webhook. |
LlmCall
Stores metadata of the call of an LLM.
JSON representation |
---|
{ "retrievedExamples": [ { object ( |
Fields | |
---|---|
retrievedExamples[] |
A list of relevant examples used for the LLM prompt. |
tokenCount |
The token counts of the LLM call. |
model |
The model of the LLM call. |
temperature |
The temperature of the LLM call. |
RetrievedExample
Relevant example used for the LLM prompt.
JSON representation |
---|
{
"exampleId": string,
"exampleDisplayName": string,
"retrievalStrategy": enum ( |
Fields | |
---|---|
exampleId |
The id of the example. |
exampleDisplayName |
The display name of the example. |
retrievalStrategy |
Retrieval strategy of the example. |
matchedRetrievalLabel |
Optional. The matched retrieval label of this LLM call. |
RetrievalStrategy
Retrieval strategy on how the example is selected to be fed to the prompt.
Enums | |
---|---|
RETRIEVAL_STRATEGY_UNSPECIFIED |
Not specified. DEFAULT will be used. |
DEFAULT |
Default retrieval strategy. |
STATIC |
Static example will always be inserted to the prompt. |
NEVER |
Example will never be inserted into the prompt. |
TokenCount
Stores token counts of the LLM call.
JSON representation |
---|
{ "totalInputTokenCount": string, "conversationContextTokenCount": string, "exampleTokenCount": string, "totalOutputTokenCount": string } |
Fields | |
---|---|
totalInputTokenCount |
The total number of tokens used for the input to the LLM call. |
conversationContextTokenCount |
The number of tokens used for the conversation history in the prompt. |
exampleTokenCount |
The number of tokens used for the retrieved examples in the prompt. |
totalOutputTokenCount |
The total number of tokens used for the output of the LLM call. |
IntentMatch
Stores metadata of the intent match action.
JSON representation |
---|
{
"matchedIntents": [
{
object ( |
Fields | |
---|---|
matchedIntents[] |
The matched intent. |
MatchedIntent
Stores the matched intent, which is the result of the intent match action.
JSON representation |
---|
{ "intentId": string, "displayName": string, "score": number, "generativeFallback": { object } } |
Fields | |
---|---|
intentId |
The ID of the matched intent. |
displayName |
The display name of the matched intent. |
score |
The score of the matched intent. |
generativeFallback |
The generative fallback response of the matched intent. |
FlowStateUpdate
Stores metadata of the state update action, such as a state machine execution in flows.
JSON representation |
---|
{ "eventType": string, "pageState": { object ( |
Fields | |
---|---|
eventType |
The type of the event that triggered the state update. |
pageState |
The updated page and flow state. |
updatedParameters |
The updated parameters. |
destination |
The destination of the transition. Format: |
functionCall |
The function call to execute. |
PageState
Stores the state of a page and its flow.
JSON representation |
---|
{ "page": string, "displayName": string, "status": string } |
Fields | |
---|---|
page |
The ID of the page. Format: |
displayName |
The display name of the page. |
status |
The status of the page. |
FunctionCall
Stores the metadata of a function call to execute.
JSON representation |
---|
{ "name": string } |
Fields | |
---|---|
name |
The name of the function call. |
PlaybookInvocation
Stores metadata of the invocation of a child playbook. Playbook invocation actions enter the child playbook.
JSON representation |
---|
{ "playbook": string, "displayName": string, "playbookInput": { object ( |
Fields | |
---|---|
playbook |
Required. The unique identifier of the playbook. Format: |
displayName |
Output only. The display name of the playbook. |
playbookInput |
Optional. Input of the child playbook invocation. |
playbookOutput |
Optional. Output of the child playbook invocation. |
playbookState |
Required. Playbook invocation's output state. |
OutputState
Output state.
Enums | |
---|---|
OUTPUT_STATE_UNSPECIFIED |
Unspecified output. |
OUTPUT_STATE_OK |
Succeeded. |
OUTPUT_STATE_CANCELLED |
Cancelled. |
OUTPUT_STATE_FAILED |
Failed. |
OUTPUT_STATE_ESCALATED |
Escalated. |
OUTPUT_STATE_PENDING |
Pending. |
FlowInvocation
Stores metadata of the invocation of a child CX flow. Flow invocation actions enter the child flow.
JSON representation |
---|
{
"flow": string,
"displayName": string,
"inputActionParameters": {
object
},
"outputActionParameters": {
object
},
"flowState": enum ( |
Fields | |
---|---|
flow |
Required. The unique identifier of the flow. Format: |
displayName |
Output only. The display name of the flow. |
inputActionParameters |
Optional. A list of input parameters for the flow. |
outputActionParameters |
Optional. A list of output parameters generated by the flow invocation. |
flowState |
Required. Flow invocation's output state. |
PlaybookTransition
Stores metadata of the transition to another target playbook. Playbook transition actions exit the caller playbook and enter the target playbook.
JSON representation |
---|
{ "playbook": string, "displayName": string, "inputActionParameters": { object } } |
Fields | |
---|---|
playbook |
Required. The unique identifier of the playbook. Format: |
displayName |
Output only. The display name of the playbook. |
inputActionParameters |
A list of input parameters for the action. |
FlowTransition
Stores metadata of the transition to a target CX flow. Flow transition actions exit the caller playbook and enter the child flow.
JSON representation |
---|
{ "flow": string, "displayName": string, "inputActionParameters": { object } } |
Fields | |
---|---|
flow |
Required. The unique identifier of the flow. Format: |
displayName |
Output only. The display name of the flow. |
inputActionParameters |
A list of input parameters for the action. |
TTS
This type has no fields.
Stores metadata of the Text-to-Speech action.
STT
This type has no fields.
Stores metadata of the Speech-to-Text action.
Span
A span represents a sub execution step of an action.
JSON representation |
---|
{
"name": string,
"tags": [
string
],
"metrics": [
{
object ( |
Fields | |
---|---|
name |
The name of the span. |
tags[] |
The metadata tags of the span such as span type. |
metrics[] |
The unordered collection of metrics in this span. |
startTime |
Timestamp of the start of the span. Uses RFC 3339, where generated output will always be Z-normalized and uses 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
completeTime |
Timestamp of the completion of the span. Uses RFC 3339, where generated output will always be Z-normalized and uses 0, 3, 6 or 9 fractional digits. Offsets other than "Z" are also accepted. Examples: |
NamedMetric
A named metric is a metric with name, value and unit.
JSON representation |
---|
{ "name": string, "value": value, "unit": string } |
Fields | |
---|---|
name |
The name of the metric. |
value |
The value of the metric. |
unit |
The unit in which this metric is reported. Follows The Unified Code for Units of Measure standard. |
Status
The status of the action.
JSON representation |
---|
{
"exception": {
object ( |
Fields | |
---|---|
exception |
Optional. The exception thrown during the execution of the action. |
Exception
Exception thrown during the execution of an action.
JSON representation |
---|
{ "errorMessage": string } |
Fields | |
---|---|
errorMessage |
Optional. The error message. |
Methods |
|
---|---|
|
Creates an example in the specified playbook. |
|
Deletes the specified example. |
|
Retrieves the specified example. |
|
Returns a list of examples in the specified playbook. |
|
Update the specified example. |