# Stagehand

> Automate browser sessions with Stagehand navigation, observation and extraction.

App · 8 actions. Web page: https://docs.lodol.com/docs/integrations/stagehand

## Actions

### Act

Act.

**Inputs**

- Session id (`session_id`, text, required): Session id.
- Instruction (`instruction`, text, required): Instruction.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {}
}
```

### End session

End session.

**Inputs**

- Session id (`session_id`, text, required): Session id.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {}
}
```

### Extract

Extract.

**Inputs**

- Session id (`session_id`, text, required): Session id.
- Instruction (`instruction`, text): Instruction.
- Schema (`schema`, object): Schema.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {}
}
```

### Get replay

Get replay.

**Inputs**

- Session id (`session_id`, text, required): Session id.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {}
}
```

### Navigate

Navigate.

**Inputs**

- Session id (`session_id`, text, required): Session id.
- Url (`url`, text, required): Url.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {}
}
```

### Observe

Observe.

**Inputs**

- Session id (`session_id`, text, required): Session id.
- Instruction (`instruction`, text): Instruction.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {}
}
```

### Run agent

Run agent.

**Inputs**

- Session id (`session_id`, text, required): Session id.
- Agent config (`agent_config`, object, required): Agent config.
- Execute options (`execute_options`, object, required): Execute options.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {}
}
```

### Start session

Start session.

**Inputs**

- Model name (`model_name`, text, required): Model identifier supported by your model provider, for example openai/gpt-4.1.
- System prompt (`system_prompt`, text): System prompt.
- Self heal (`self_heal`, true/false): Self heal.
- Browserbase session id (`browserbase_session_id`, text): Browserbase session id.

**Outputs**

- `data` (object): Service response.

**Example output**

```json
{
  "data": {
    "data": {
      "available": false,
      "sessionId": "example"
    },
    "success": false
  }
}
```
