> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bubench.lexmount.io/llms.txt
> Use this file to discover all available pages before exploring further.

# cli_utils

> CLI argument parsing utilities

# browseruse\_bench.utils.cli\_utils

Command-line argument parsing and error handling utilities.

## Import

```python theme={null}
from browseruse_bench.utils import (
    handle_cli_errors,
    create_run_parser,
    create_eval_parser,
    create_base_agent_parser,
)
```

***

## handle\_cli\_errors

CLI error handling decorator for unified exception and exit code handling.

```python theme={null}
def handle_cli_errors(func: Callable[..., int]) -> Callable[..., None]
```

### Usage Example

```python theme={null}
from browseruse_bench.utils import handle_cli_errors

@handle_cli_errors
def main() -> int:
    # Your logic
    return 0  # Success

if __name__ == "__main__":
    main()
```

### Behavior

| Condition         | Exit Code             |
| ----------------- | --------------------- |
| Normal return     | Function return value |
| KeyboardInterrupt | 130                   |
| Other exceptions  | 1                     |

***

## create\_run\_parser

Create argument parser for the main run script.

```python theme={null}
def create_run_parser() -> argparse.ArgumentParser
```

### Included Arguments

| Argument           | Type   | Default | Description                                                            |
| ------------------ | ------ | ------- | ---------------------------------------------------------------------- |
| `--mode`           | choice | `all`   | Test mode: `single`, `first_n`, `specific`, `sample_n`, `by_id`, `all` |
| `--count`          | int    | 1       | Number of tasks for `first_n` or `sample_n` mode                       |
| `--task-ids`       | list   | -       | Task ID list for `specific` mode                                       |
| `--id`             | str    | -       | Single task ID for `by_id` mode                                        |
| `--timeout`        | int    | -       | Timeout per task (seconds)                                             |
| `--skip-completed` | flag   | False   | Skip completed tasks                                                   |
| `--dry-run`        | flag   | False   | Show command only, do not execute                                      |

***

## create\_eval\_parser

Create argument parser for the evaluation script.

```python theme={null}
def create_eval_parser() -> argparse.ArgumentParser
```

### Included Arguments

| Argument            | Type | Default | Description                                     |
| ------------------- | ---- | ------- | ----------------------------------------------- |
| `--mode`            | str  | -       | Evaluation mode                                 |
| `--model`           | str  | -       | Evaluation model                                |
| `--score-threshold` | int  | -       | Score threshold (LexBench-Browser=60, others=3) |
| `--num-worker`      | int  | 1       | Number of worker processes                      |
| `--api-key`         | str  | -       | API Key                                         |
| `--base-url`        | str  | -       | API Base URL                                    |
| `--dry-run`         | flag | False   | Show command only, do not execute               |

***

## create\_base\_agent\_parser

Create base argument parser for Agent run.py.

```python theme={null}
def create_base_agent_parser(
    description: str,
    default_tasks_json: str,
    default_output_dir: str
) -> argparse.ArgumentParser
```

<ParamField path="description" type="str" required>
  Parser description
</ParamField>

<ParamField path="default_tasks_json" type="str" required>
  Default tasks JSON file path
</ParamField>

<ParamField path="default_output_dir" type="str" required>
  Default output directory
</ParamField>

### Usage Example

```python theme={null}
from browseruse_bench.utils import create_base_agent_parser

parser = create_base_agent_parser(
    description="Browser-Use Agent Runner",
    default_tasks_json="data/tasks.json",
    default_output_dir="output/"
)
args = parser.parse_args()
```