Member Junction
    Preparing search index...

    Interface DelegatedResult

    The result of a delegated target-agent run, fed back to the realtime model as a tool response.

    interface DelegatedResult {
        Artifacts?: DelegatedRunArtifact[];
        CallID: string;
        Output: string;
        PausedRunID?: string;
        RunID?: string;
        Success: boolean;
    }
    Index

    Properties

    Artifacts? CallID Output PausedRunID? RunID? Success

    Properties

    Artifacts?: DelegatedRunArtifact[]

    Artifacts the delegated run produced (when artifact creation is enabled for the target and the run returned a payload). Serialized into the tool result as artifacts so the call overlay can auto-open them in artifact tabs. Absent/empty when the run produced none.

    CallID: string

    The provider call ID this result corresponds to.

    Output: string

    The textual outcome to narrate (on success) or the error to surface (on failure). The driver feeds this back to the model as the tool_response so it can speak the outcome.

    PausedRunID?: string

    When the delegated run paused awaiting user feedback (an interactive target agent, e.g. Query Builder confirming a task graph), this carries the paused run's id so the transport layer can persist it and resume the SAME run on the user's next answer (rather than starting a fresh run). Absent when the run completed (or failed) without pausing. The Output in this case is the agent's clarifying QUESTION, and Success is true (a valid intermediate outcome).

    RunID?: string

    ID of the delegated target-agent run (MJ: AI Agent Runs), when one was created. Serialized into the tool result as runId so client developer tooling (the call overlay's dev links) can open the underlying run record. Absent when the delegation failed before a run existed.

    Success: boolean

    Whether the delegated run completed successfully.

    Settings

    Member Visibility

    On This Page

    Properties