core/models/index library
Central export file for all Tuul AI models.
This barrel file allows you to access all core data structures, configuration options, and response models using a single import:
import 'package:tuul_ai/core/models/index.dart';
Comprehensive configuration for standard AI generation.
Classes
- AbilityOptions
- Toggles specific features of the AI model.
-
ActionResponse<
T> - Represents a response that requires further action from the client.
- AgentOptions
- Identifies the specific AI Agent configuration to be used.
-
AppResponse<
T> - A sealed base class representing all possible outcomes from a Tuul AI API request.
- BundleRunResponse
- A comprehensive response object returned by a standard generation run.
- ChainOfThoughtStepType
- A single discrete step within a Chain of Thought.
- ChainOfThoughtType
- A collection of steps representing the AI's "Chain of Thought."
- CheckpointType
- A marker indicating the state of the conversation at a specific point in time.
- ConfirmationType
- Represents the approval state for an action that requires user consent.
- ContextDataType
- Contextual metadata including the model used and detailed token usage.
- Conversation
- A comprehensive model representing a Tuul AI conversation session and its metadata.
- ConversationError
- Represents an error entry within a conversation's lifecycle.
- ConversationHistoryResponse
- Represents a paginated or grouped response containing conversation summaries.
-
ErrorResponse<
T> - Represents a failed API call or a business logic error from the server.
- ErrorSseEvent
- Emitted when a terminal or non-terminal error occurs within the stream.
- FunctionCall
- Represents a specific request from the AI model to execute an external function.
- FunctionCallResponse
- The response object returned after executing an AI function call.
- FunctionDeclaration
- A collection of tools/functions declared to the AI model.
- FunctionDeclarationData
- Details about the retrieved function declarations.
- ImageOptions
- Metadata for attachments (images/documents) included in a prompt.
- LifecycleSseEvent
- Represents lifecycle transitions in the stream (e.g., 'start', 'finish', 'done').
- LiteCharacteristic
- A key-value pair used to define specific behavioral traits for the AI.
-
LiteGenHistoryResponse<
T> - Represents a structured response containing a history of lite generation messages.
- LiteGenPart
- A atomic unit of content within a LiteMessage.
- LiteMessage
- A simplified message model used in lite generation.
- LiteOption
- Configuration options for a LiteGen request.
- MessageResponse
- Represents a response containing the detailed message history of a specific conversation.
- MessageVersionType
- Represents a specific historical iteration of a message's content.
- Parameters
- Parameters for a tool, following JSON Schema standards.
- PlanType
- The high-level plan the AI generated before executing its response.
- PromptOptions
- Contains the primary input data for the AI.
- Properties
- A simple wrapper for tool property maps.
- RawSseEvent
- A generic container for file, source, or raw data events.
- ReasoningDeltaSseEvent
- Emitted when a new chunk of the AI's internal reasoning is available.
- ReasoningType
- Contains the model's reasoning content and processing time.
- RegularResponse
- A wrapper for standard, non-streaming asynchronous responses.
- RichMessage
- The primary data model for advanced AI interactions.
- SessionOptions
- Contextual IDs used to track long-running conversations.
- SimpleMessageType
- A simplified message structure used within the response queue.
- SourceType
- Metadata for a source citation used in a grounded AI response.
- SseEvent
- A sealed base class representing all possible events in a Tuul AI stream.
- StateOptions
- Controls the lifecycle of the conversation state.
- StepSseEvent
- Represents the beginning or end of a discrete logical step in a multi-step process.
- StreamResponse
- A wrapper for streaming responses, typically used in generative AI scenarios.
-
SuccessResponse<
T> - Represents a successful API call.
- TaskType
- A key-value pair representing a task identified by the AI.
- TextDeltaSseEvent
- Emitted when a new chunk of generated text is available.
- ToggleStatusResponse
- Response returned when toggling the active state of a function or tool.
- ToggleStatusResponseData
- Data payload for status toggle events.
- Tool
- A specific tool definition including its parameters.
- ToolCallSseEvent
- Emitted when the AI decides to call a specific tool or function.
- ToolCallType
- Details of an AI tool or function call.
- ToolErrorSseEvent
- Emitted if a tool execution fails.
- ToolInputDeltaSseEvent
- Emitted as the AI streams the JSON input arguments for a tool call.
- ToolResultSseEvent
- Emitted when a tool has successfully finished execution and returned data.
- TriggerLog
- Log entry generated during a trigger execution.
- TriggerParse
- A container class that holds the parsed result of a LiveFunction interaction.
-
TriggerResponse<
T> - A generic response wrapper for trigger-based interactions.
- TuulOptions
- The master configuration class for Tuul AI requests.
- UnknownSseEvent
- A fallback event type for any server events not explicitly handled by the SDK version.
- Usage
- Detailed token usage statistics for a single request/response cycle.
Typedefs
- FunctionCallResult = dynamic
- Represents the raw output of a function call.
- SuggestionType = String
- A string-based suggestion for a user's next message.