Fix TypeError in aa_omniscience_prompt

## Describe the bug
The `aa_omniscience` prompt function was missing the `task_name` argument in its signature or call, causing a `TypeError` when the evaluator attempted to pass the task name.

## To Reproduce
```python
task = "aa_omniscience|5"

pipeline = Pipeline(
tasks=task,
pipeline_parameters=pipeline_params,
evaluation_tracker=evaluation_tracker,
model_config=model_config,
)

pipeline.evaluate()
pipeline.save_and_push_results()
pipeline.show_results()
```
```
140 # We init tasks first to fail fast if one is badly defined
141 self._init_random_seeds()
--> 142 self._init_tasks_and_requests(tasks=tasks)
144 self.model_config = model_config
145 self.accelerator, self.parallel_context = self._init_parallelism_manager()
...
--> 309 doc = self.formatter(item, self.name)
310 # Skip if formatter returns None (e.g., to filter out certain samples)
311 if doc is None or doc == []:

TypeError: aa_omniscience_prompt() takes 1 positional argument but 2 were given
```

## Expected behavior
The function should accept `task_name` as an optional argument and return a `Doc` object containing that `task_name`.

## Version info
- OS: mac
- Lighteval version: main (local development)