How Hero AI Executes Components

hero ai enables users to trigger component based automation directly from natural language prompts in the hero ai companion this functionality is available for components that are marked as visible to hero ai in the component builder component discovery when a user sends a prompt in the hero ai companion (for example, “send slack message to the networking team”), hero ai interprets the intent and searches for matching components based on the component's name and description input field names and expected schema associated interfaces or user defined configuration when a match is found, hero ai displays a component card with a message prompting the user to review the inputs and confirm execution review and edit inputs click review details to inspect the component's input fields if needed, you can modify the values pre filled by hero ai these values may be inferred from the current workspace or record context user profile details defaults defined in the component you can adjust values for strings, numbers, booleans, arrays, and object structures click save to confirm your changes execution after reviewing the details, click run to execute the component hero ai sends the inputs to the component and triggers execution the result will appear in the chat window, confirming that the automation completed if the component is configured for auto execution ( requires confirmation to execute is disabled), the component runs immediately after a match is found—no review or approval is needed note if hero ai is unable to resolve required inputs from context, it will prompt the user to manually provide those values before continuing which components are best suited for hero ai? as a best practice, we generally recommend making simple, modular and straight forward turbine components visible to hero ai components that require complex nested json objects as input may be more difficult for hero ai to populate component execution current limitations rbac components marked as " visible to hero al " will be available to be called by all users in the tenant there are no current permissions we can configure to specify which users can use which components component runtime errors in scenarios where component executions result in errors, hero al in some instances will show a generic error without a lot of error details users without access to orchestration features won't be able to independently troubleshoot and view