Measured

class gl:
    clear = lambda a: a * 3
print(gl.clear(2))          # CPython 6; pxx TypeError: object is not callable

The control that says where it lives:

f = lambda a: a * 3
print(f(2))                 # 6 in both

And the same shape through the new staticmethod arm behaves identically — clear = staticmethod(lambda a: a * 3) also raises — which is what establishes that the wrapper is not involved. A bare def name in the same position works (clear = add_one then gl.clear(1) answers 2), so the difference is specifically the lambda's boxed closure value reaching a class attribute.

Why it compiles

PyBoxCallableValue is what makes a lambda callable through a name, and the module-level assignment path calls it. The class-attribute store is a different route. Two paths for one concept, and the one nobody extended is the one that is broken — normalise-dont-special-case.

What a fix must carry

Both doors, since they are separate: gl.clear(2) through the CLASS and gl().clear(2) through an INSTANCE. Note the instance door has its own open divergence for plain functions ([[bug-n-a-plain-function-as-a-class-attribute-does-not-bind-the-receiver]]), so a lambda row there needs to say which behaviour it is asserting.

Fixed 2026-10-02

PyEmitClassAttrExpr stored the lambda's closure pointer into the hidden class-attribute global raw; the module-level g = lambda path boxes it with PyBoxCallableValue, and the class-body path now does the same. While there, type() of any def, lambda or bound method printed <class 'NoneType'> (no builtin type code carried 'function' or 'method'); PYBT_FUNCTION and PYBT_METHOD close that. Test: test/test_nilpy_a_lambda_in_a_class_attribute_is_a_callable_function.npy (HEAP_DEBUG and i386 rows, diffed against CPython).