Runtime 与运行结果

Runtime 组合模型、工具、Skills 与会话。使用 async with 初始化并释放资源，复用同一 Runtime 可以发起多次运行。

快速开始 · 流式输出 · 会话与历史

常用接口

接口	用途
from_config	从配置对象、字典或 JSON 路径创建 Runtime。
BumblehiveRuntime	run() 获取结果；stream() 消费事件；run_console() 在终端展示过程。history 与 session_id 不能同时使用，approval_handler 仅影响当前调用。
AgentRunResult	final_content 是最终回答，usage 是模型用量，stop_reason 说明结束原因。首先检查 error；tools_used 只记录成功执行的工具。

创建与运行

从配置对象、字典或 JSON 路径创建 Runtime。

from_config

from_config(config: ConfigInput = None) -> BumblehiveRuntime

Create a runtime from a JSON file, mapping, or config object.

run() 获取结果；stream() 消费事件；run_console() 在终端展示过程。history 与 session_id 不能同时使用，approval_handler 仅影响当前调用。

BumblehiveRuntime

High-level agent runtime built from BumblehiveConfig.

from_config classmethod

from_config(config: ConfigInput = None) -> BumblehiveRuntime

Create a runtime from a JSON file, mapping, config object, or defaults.

initialize_tools async

initialize_tools() -> None

Register built-ins and connect the runtime-scoped MCP servers once.

run async

run(message: UserMessage, *, config: Mapping[str, Any] | None = None, hooks: HookInput = None, history: MessageHistory | None = None, session_id: str | None = None, approval_handler: ToolApprovalHandler | None = None) -> AgentRunResult

Run one conversation turn.

With neither history nor session_id, the turn is stateless. Passing history reads caller-owned in-memory history without updating it. Use history.replace_run_messages(result.messages) to carry a completed run into the next turn. Passing session_id loads, recovers, and persists a managed session. Passing both raises ValueError.

Do not share one MessageHistory across concurrent runs: await turns sequentially, or use separate histories for parallel conversations.

stream

stream(message: UserMessage, *, config: Mapping[str, Any] | None = None, hooks: HookInput = None, history: MessageHistory | None = None, session_id: str | None = None, approval_handler: ToolApprovalHandler | None = None, max_queue_size: int = DEFAULT_STREAM_QUEUE_SIZE) -> AsyncEventStream[AgentRunResult]

Stream one turn with stateless, in-memory, or persisted history.

run_console async

run_console(message: UserMessage, *, config: Mapping[str, Any] | None = None, hooks: HookInput = None, history: MessageHistory | None = None, session_id: str | None = None, approval_handler: ToolApprovalHandler | None = None, renderer: Any | None = None, max_queue_size: int = DEFAULT_STREAM_QUEUE_SIZE) -> AgentRunResult

Run and render one turn with optional conversation history.

delete_session async

delete_session(session_id: str) -> bool

Delete a persisted session and evict its cached state.

close async

close() -> None

Release resources owned by this runtime.

读取结果

final_content 是最终回答，usage 是模型用量，stop_reason 说明结束原因。首先检查 error；tools_used 只记录成功执行的工具。

AgentRunResult dataclass

Outcome of one tool-capable model run.