Preview an AI action against a sample item
POST
/api/v1/ai/actions/{actionId}/preview
const url = 'https://app.avera.io/api/v1/ai/actions/example/preview';const options = { method: 'POST', headers: {'Content-Type': 'application/json'}, body: '{"sampleSourceItemId":"2489E9AD-2EE2-8E00-8EC9-32D5F69181C0"}'};
try { const response = await fetch(url, options); const data = await response.json(); console.log(data);} catch (error) { console.error(error);}curl --request POST \ --url https://app.avera.io/api/v1/ai/actions/example/preview \ --header 'Content-Type: application/json' \ --data '{ "sampleSourceItemId": "2489E9AD-2EE2-8E00-8EC9-32D5F69181C0" }'A dry run: assembles the context and invokes the model against a sample source item, returning the transient result without persisting a proposal. The call still spends tokens on the tenant’s own key (BYOK) and decrements the rate-limit budget, and every call is audit-logged.
Parameters
Section titled “Parameters”Path Parameters
Section titled “Path Parameters”actionId
required
string
Request Bodyrequired
Section titled “Request Bodyrequired”Media typeapplication/json
object
sampleSourceItemId
required
string format: uuid
Example generated
{ "sampleSourceItemId": "2489E9AD-2EE2-8E00-8EC9-32D5F69181C0"}Responses
Section titled “Responses”OK
Media type*/*
object
actionId
required
string
assembledSystemPrompt
required
string
assembledUserPrompt
required
string
slots
required
object
key
additional properties
string
generatedContent
required
string
confidenceScore
required
number format: float
confidenceDetails
required
object
key
additional properties
number format: float
modelId
required
string
inputTokens
required
integer format: int32
outputTokens
required
integer format: int32
totalTokens
required
integer format: int32
estimatedCostUsd
required
number format: double