promptAgentActivity
Examples
type PromptAgentActivity implements AgentActivity {
id: ID!
createdAt: DateTime
activityType: AgentActivityType!
oxAction: OxAction!
userEmail: String
userRole: String
source: VibeCodeSource!
appName: String
threadId: String
promptSizeBytes: Int
responseTokenCount: Int
model: String
appliedGuidelines: [AppliedGuideline!]
secretRuleIds: [String!]
maskedSecrets: [String!]
secretTypeLabels: [String!]
blockedSecretCount: Int
}Fields
id ID!
Unique identifier of the activity record
createdAt DateTime
When the activity was recorded
userEmail String
Email address of the developer who triggered the activity
userRole String
Role of the developer who triggered the activity
appName String
Application the activity was attributed to
threadId String
Identifier of the agent conversation the activity belongs to
promptSizeBytes Int
Size of the prompt, in bytes
responseTokenCount Int
Tokens of security guidance the model received for this prompt, as the hook injected it, wrapper text included
model String
AI model the prompt was sent to
appliedGuidelines [AppliedGuideline!]
OX guidelines injected into this prompt
oxGuidelineId String!
displayName String!
category String!
createdBy String!
prompt String!
promptSizeBytes Int
isFromCache Boolean
tokenCount Int
secretRuleIds [String!]
Identifiers of the secret-detection rules that matched
maskedSecrets [String!]
Matched credentials, already masked by the hook
secretTypeLabels [String!]
Human-readable names of the secret types that matched (e.g. "AWS Access Token"), derived from secretRuleIds. Never derived from the credential itself
blockedSecretCount Int
How many secret matches the hook reported (capped at 10 by the hook)
References
Queries using this object:
Last updated