_.isInt

_.isInt answers a deceptively simple question - "is this object an integer?" - with a deceptively strict implementation. It's type(self.obj) is int, the exact-type check, from the sfinktah/comfy-ovum pack's ovum/underscore family (auto-generated wrappers around underscore3, a Python port of Underscore.js). The strictness is the whole story here, because the thing that trips people up isn't the node - it's what counts as an int in a ComfyUI graph.

The family's README disclaimer applies: work in progress, "not recommended for use in workflows." Fair enough as a general caution. This particular node is trivial, though, and the behavior is fully predictable if you know one Python fact: type(x) is int is exact.

What it does

The check is type(self.obj) is int. Work through what that means:

• A plain Python int (widget values, INT-typed node outputs) → True.
• A float like 3.0 → False (it's a float, not an int, even though it's mathematically integral).
• A bool → False. In Python, True and False are instances of int, but their type is bool, so the exact check says no. If you want bools to count, use _.isIntegral.
• A NumPy scalar like numpy.int64(5) → False. This is the one that bites. Lots of ComfyUI custom nodes compute values with numpy or torch and hand back numpy scalars, which are not the Python int type. A value that looks like 5 and acts like 5 reads as "not an int" here.

If your numbers only come from Comfy widgets, _.isInt works exactly as you'd expect. If they come from computed paths, expect surprises.

Like the rest of the family, it takes a plain value and returns the boolean, or takes a _.CHAIN from _.chain and keeps the chain going.

Inputs and outputs

• obj - the only input, type *, optional.
• Output - a single * socket holding a Python bool (family quirk: value is boolean, socket is typed generic).

Where it fits

Honestly, this one's for validation and branching rather than day-to-day use. A typical case: a dynamic node returns a value that could be a string or a number, and you want to route it differently based on whether it's a genuine Python int. Or you're debugging why a numeric input isn't behaving and want to check the actual type coming off a wire. For "is this a whole number, roughly" questions, _.isIntegral is the better tool; for "is this literally a Python int," this is it.

Installing it

Install the pack via ComfyUI Manager - search comfy-ovum - or:

cd ComfyUI/custom_nodes
git clone https://github.com/sfinktah/comfy-ovum

Restart ComfyUI. underscore3 is vendored in the repo, so no extra install for this node.

Gotchas

• numpy.int64 and torch tensor scalars read as False even though they're integer-valued. If you're checking computed values, that's probably not what you wanted.
• Bools read as False - True is technically an int in Python, but not an exact int type.
• 3.0 reads as False; it's a float. Use _.isIntegral if you want "mathematically integral" to count.
• The output wire is *, not BOOLEAN.