# RespondDecision

> **Class** in `langchain`

📖 [View in docs](https://reference.langchain.com/python/langchain/agents/middleware/human_in_the_loop/RespondDecision)

Response when a human answers on behalf of the tool, skipping execution.

Used for "ask user" style tools whose real implementation is the human's
response. The tool is not executed; instead, a synthetic `ToolMessage` with
`status="success"` and the provided `message` is returned to the model.

## Signature

```python
RespondDecision()
```

## Extends

- `TypedDict`

## Constructors

```python
__init__(
    type: Literal['respond'],
    message: str,
)
```

| Name | Type |
|------|------|
| `type` | `Literal['respond']` |
| `message` | `str` |


## Properties

- `type`
- `message`

---

[View source on GitHub](https://github.com/langchain-ai/langchain/blob/85a5a04210452aec2eb3a06a02961c8fefd5c8b6/libs/langchain_v1/langchain/agents/middleware/human_in_the_loop.py#L98)