Autousing fixtures while testing fns which have explicit type-checks?

[xor-xor]

Let's say that I need to test some function, which explicitly checks the type of its input, and I have no control over this behavior. Also, let's say, that I need a fixture for such test:

import pytest


@pytest.fixture()
def my_fixture():
    return 1


def alien_fun(x):
    assert type(x) == int, "sorry, integers only!"
    return x + 1


def test_alien_fun(my_fixture):
    print(type(my_fixture))

    result = alien_fun(my_fixture)

    assert result == 2

So far, so good - everything works as expected and I get <class 'int'> in the output.

BUT!

If I change this fixture to autouse=True (because e.g. I have a lots of tests like the one above and I just don't want to repeat the same declaration over and over):

import pytest


@pytest.fixture(autouse=True)
def my_fixture():
    return 1


def alien_fun(x):
    assert type(x) == int, "sorry, integers only!"
    return x + 1


def test_alien_fun():
    print(type(my_fixture))

    result = alien_fun(my_fixture)

    assert result == 2

...then everything breaks with AssertionError: sorry, integers only! (well, obviously), because my_fixture is now <class 'function'>.

Assuming that this is a feature, not a bug, my question is: what is the "canonical" way of handling such situations?

EDIT: actually, I've just realized that this type assertion here is irrelevant - this autouse=True case will just "explode" on return x + 1 for exactly the same reason.