LlmStartPayload

Payload for “llm_start” events — an LLM request is about to be sent.

Class Diagram

---
title: LlmStartPayload
config:
  look: handDrawn
  theme: colorful
  class:
    hideEmptyMembersBox: true
---
classDiagram
    class LlmStartPayload {
        +string provider
        +string modelId
        +int32 messageCount
        +int32 attempt
    }

Yaml Example

provider: openai
modelId: gpt-4o-mini
messageCount: 4
attempt: 0

Properties

Name	Type	Description
provider	string	Provider identifier used for the request
modelId	string	Model or deployment identifier used for the request
messageCount	int32	Number of messages sent to the provider
attempt	int32	Retry attempt number, zero for the initial attempt