dragon.ai.agent.config.pipeline.TaskResult

class TaskResult[source] 

Bases: object

Lightweight token passed between Dragon Batch node functions.

This object is never written to the distributed dictionary. It flows purely as a Python return value between dispatcher closures executed by Dragon Batch.

serialized_ddict carries the shared DDict handle so plain-function nodes can attach directly via upstream.serialized_ddict without needing it injected as a separate argument.

Parameters

task_id:: Unique identifier for the pipeline task.
agent_id:: Identifier of the agent that produced this result.
status:: Canonical completion status (coerced to TaskStatus).
serialized_ddict:: Serialized handle of the shared DDict for the current run.
metadata:: Arbitrary extra data attached by the agent or dispatcher.

__init__(task_id: str , agent_id: str , status: TaskStatus, serialized_ddict: str = '', metadata: dict[str, ~typing.Any]=<factory>) → None 

Methods

__init__(task_id, agent_id, status, ...)

Attributes

`serialized_ddict`
`task_id`
`agent_id`
`status`
`metadata`

task_id: str 

agent_id: str 

status: TaskStatus

serialized_ddict: str = ''

metadata: dict [str , Any ]

__init__(task_id: str , agent_id: str , status: TaskStatus, serialized_ddict: str = '', metadata: dict[str, ~typing.Any]=<factory>) → None 