Class OpenAIRunnerFactory

Factory for creating OpenAI runners (chat completion and agent).

A single factory shares one OpenAI client across all runners it produces so connection pooling and instrumentation are preserved.

Hierarchy

AIProvider
- OpenAIRunnerFactory

Index

Constructors

Properties

_client _logger?

Methods

createAgent createAgentGraph createModel getClient _ensureInstrumented

Constructors

constructor

new OpenAIRunnerFactory(logger?): OpenAIRunnerFactory
Parameters
- Optional logger: LDLogger
  Optional
Returns OpenAIRunnerFactory
Overrides AIProvider.constructor
- Defined in ai-providers/server-ai-openai/src/OpenAIRunnerFactory.ts:21

Properties

`Private` _client

_client: OpenAI

`Protected` `Optional` _logger

_logger?: LDLogger

Methods

createAgent

createAgent(config, tools?): Promise<OpenAIAgentRunner>
Create an agent runner from an agent AI configuration.

The returned runner uses the OpenAI Agents SDK (@openai/agents) which manages its own OpenAI client internally.
Parameters
- config: LDAIAgentConfig
  
  The LaunchDarkly AI agent configuration. Tool definitions are sourced from config.tools.
- Optional tools: ToolRegistry
  
  Registry mapping tool names to their callable implementations or pre-built openai-agents tool instances. Tool names referenced by the config that are not present here will be logged and skipped.
  
  Optional
Returns Promise<OpenAIAgentRunner>
Overrides AIProvider.createAgent
- Defined in ai-providers/server-ai-openai/src/OpenAIRunnerFactory.ts:56

createAgentGraph

createAgentGraph(_graphDef, _tools?): Promise<undefined | AgentGraphRunner>
Create an AgentGraphRunner for an agent graph definition.

Override in provider subclasses to return a configured AgentGraphRunner. Default implementation returns undefined.
Parameters
- _graphDef: AgentGraphDefinition
- Optional _tools: ToolRegistry
  Optional
Returns Promise<undefined | AgentGraphRunner>
Promise resolving to an AgentGraphRunner, or undefined if this provider does not support graph execution.
Inherited from AIProvider.createAgentGraph
- Defined in sdk/server-ai/dist/index.d.ts:1236

createModel

createModel(config, multiTurn?): Promise<OpenAIModelRunner>
Create a model runner from a completion AI configuration.
Parameters
- config: LDAICompletionConfig
  
  The completion (or judge) AI configuration.
- multiTurn: boolean = true
  
  Whether the runner should accumulate conversation history across successive run() calls. Defaults to true (chat semantics). Pass false for stateless runners such as judges.
Returns Promise<OpenAIModelRunner>
Overrides AIProvider.createModel
- Defined in ai-providers/server-ai-openai/src/OpenAIRunnerFactory.ts:37

getClient

getClient(): OpenAI
Get the underlying OpenAI client instance.

Returns OpenAI
- Defined in ai-providers/server-ai-openai/src/OpenAIRunnerFactory.ts:98

`Static` `Private` _ensureInstrumented

_ensureInstrumented(logger?): Promise<void>
Automatically patches the ESM openai module for OpenTelemetry tracing when a TracerProvider is active and @traceloop/instrumentation-openai is installed.
Parameters
- Optional logger: LDLogger
  Optional
Returns Promise<void>
- Defined in ai-providers/server-ai-openai/src/OpenAIRunnerFactory.ts:106

Generated using TypeDoc