5 Python Anti-Patterns Every Code Reviewer Catches

Python is forgiving until it isn't. These five patterns are the ones that survive linters but get caught by every code reviewer — human or AI.

1. Mutable default arguments

def add_item(item, items=[]):  # 💀
    items.append(item)
    return items

The default list is created once at definition time and shared across every call. Every caller after the first sees the previous caller's items. Use None:

def add_item(item, items=None):
    items = items if items is not None else []
    items.append(item)
    return items

2. Late binding in loops

callbacks = [lambda: print(i) for i in range(3)]
for c in callbacks:
    c()  # prints 2 2 2, not 0 1 2

Each lambda captures the variable, not the value. Bind it as a default arg:

callbacks = [lambda i=i: print(i) for i in range(3)]

3. Bare except clauses

try:
    risky_op()
except:           # catches KeyboardInterrupt, SystemExit, anything
    pass

Always specify the exception class. except Exception is barely better — name the actual error you expect.

4. Modifying a list while iterating

for item in items:
    if not item.is_valid:
        items.remove(item)  # skips items, breaks math

Build a new list with a comprehension instead: items = [i for i in items if i.is_valid].

5. Using == for None / True / False

if value == None:    # works, but slow + non-idiomatic
if value is None:    # ✅ identity check, idiomatic

Same for True and False — use is. == can be overridden by classes and produce surprising results.

How to catch these in your own code

Pylint and Ruff will catch some of these. The remaining ones — especially anti-patterns 1, 2, and 4 — are best caught by paste-into-an-AI-reviewer flow. Every one of these is in the training data; the AI flags them instantly. Try TrashMyCode on a Python file and watch how many it catches that your linter missed.