Member Junction
    Preparing search index...

    Action that executes any top-level AI Agent by name or ID.

    Serves two purposes:

    1. External callers (schedulers, workflows, MCP clients, other Runtime actions invoking via utilities.actions.Invoke) can run agents through the uniform Action catalog instead of reaching for the AgentRunner API.
    2. Gives us a concrete dispatch target for AIAgent.ExposeAsAction once the auto-registration pass lands — every exposed agent becomes callable as an action without a separate Custom class per agent.

    Not used inside the Runtime Actions bridge. Script code calling utilities.agents.Run(...) from the sandbox bridges back to the host which invokes AgentRunner.RunAgent() directly — skipping this wrapper because the bridge already runs in-process with contextUser and doesn't need the action-catalog indirection.

    await runAction({
    ActionName: 'Execute Agent',
    Params: [
    { Name: 'AgentName', Value: 'Customer Sentiment Analyzer', Type: 'Input' },
    { Name: 'ConversationMessages', Value: [{ role: 'user', content: 'Analyze ticket 12345' }], Type: 'Input' },
    { Name: 'Data', Value: { ticketId: '12345' }, Type: 'Input' }
    ]
    });

    Hierarchy (View Summary)

    Index

    Constructors

    Properties

    Methods

    Constructors

    Properties

    AuthorizesCaller: boolean

    True when the action authorizes its caller itself, so a public entry point such as the RunAction mutation may run it for an authenticated user without an MJ: Action Authorizations link. Set it only on an action that either checks an authorization the caller must hold, or reads and writes only with the caller's own permissions. Leave it false for an action that acts with more authority than its caller, or that reaches data or sends messages the caller chooses but could not otherwise reach.

    The engine reads the marker only from the class that runs, never from its parent classes. A subclass that overrides a marked action must declare the marker again.

    Methods