AI State Handler icon

AI State Handler

Intelligent state management for AI parsing and tool invocation

AI State Handler

AI-generated

Overview

The AI State Handler node manages conversational state intelligently by integrating with a user-defined state management workflow. It processes user or system messages to update or retrieve state information, invoking AI language models and tools as needed. This node is useful for scenarios where maintaining and updating conversation context is critical, such as chatbots, virtual assistants, or AI-driven workflows that require dynamic state tracking and tool invocation based on message content.

Use Case Examples

  1. A chatbot that tracks user preferences and updates them based on user messages.
  2. An AI assistant that extracts system-level information from system messages to maintain an accurate conversation state.
  3. A workflow that uses external tools to enrich or verify state data before updating the conversation context.

Properties

Name Meaning
State Management Workflow Selects the workflow responsible for handling state storage, which must support 'get' and 'set' operations with 'operation' and 'content' input fields.
Session ID Unique identifier for the conversation session to track state per session.
Role Specifies whether the message is from the user or system, affecting how the state is processed.
Message The message content to be processed for state updates or analysis.
State Model A JSON object defining the structure and descriptions of the state fields to be managed.

Output

JSON

  • state - The updated state object after processing the message.
  • prevState - The previous state before the current update.
  • stateChangedProps - List of state properties that were changed during processing.
  • toolsInvoked - Details of any tools invoked during processing, including results or errors.
  • role - The role of the message processed (user or system).
  • message - A summary message indicating whether the state was updated and which fields changed.

Dependencies

  • Requires an AI language model connection for processing messages.
  • Optionally uses AI tools connected to the node for enriching or verifying state data.
  • Requires a user-defined state management workflow with an Execute Workflow Trigger supporting 'get' and 'set' operations.

Troubleshooting

  • Ensure the State Management Workflow is selected and properly configured with 'operation' and 'content' input fields; otherwise, the node will throw an error.
  • The Session ID must be provided and unique per conversation session; missing this will cause an error.
  • The State Model must be valid JSON; invalid JSON will cause parsing errors.
  • The node requires an AI language model connection; if not connected, it will throw an error.
  • If tools are invoked, ensure they are correctly connected and implement an 'invoke' method; otherwise, tool invocation will fail.
  • Errors parsing JSON responses from the AI or tools indicate malformed or unexpected output; verify prompt templates and AI responses.

Discussion