python-anti-patterns

Use a categorized checklist covering infrastructure, architecture, error handling, resources, type safety, and testing. Compare your code against known BAD patterns like bare except blocks, hard-coded config, and blocking calls in async functions, then apply the documented fixes.

The most common mistakes are bare `except Exception: pass` blocks that silently swallow bugs, ignoring partial failures in batch processing, and missing input validation at API boundaries. Fix these by catching specific exceptions, returning BatchResult objects, and validating with Pydantic.

Blocking calls like `time.sleep()` or synchronous `requests.get()` inside async functions freeze the entire event loop. Replace them with `asyncio.sleep()` and async-native libraries like httpx.AsyncClient to restore concurrency.

No. Double retry—where both the application and the HTTP client retry—multiplies attempts and can amplify outages. Retry at exactly one layer and know your infrastructure's built-in retry behavior before adding your own.

This checklist focuses on what to avoid, not positive design guidance. For architecture and design pattern recommendations, use a dedicated design patterns resource instead. It also does not replace automated linters or type checkers.