[Feature]: Add an extract_json() custom function to the allowed functions in tasks

[HamadaSalhab]

🔖 Feature description

Example:

Consider the following output from a prompt step for example:

I have analyzed the data, and here are my findings:

```json
{
  "foo": "bar",
  "joe": "biden"
}```

Let me know if you have any other questions:

Having a utility extract_json() would make it easier for the users to extract the json content by doing something like this:

- evaluate:
    analysis: extract_json(_)

Instead the current way, which is:

- evaluate:
    analysis: load_json(_[_.find("```json")+7:][:_[_.find("```json")+7:].find("```")])

And now analysis would be a dictionary that contains "foo" and "joe" as keys, "bar" and "biden" as values respectively.

🎤 Why is this feature needed ?

No response

✌️ How do you aim to achieve this?

No response

👀 Have you searched issues and PRs to see if this feature request has been raised before?

• I checked and didn't find similar issue