python-expert

🐍 Python Expert — Production-Grade Python

Philosophy: Python's simplicity is its power. Write code that reads like well-edited prose — explicit, flat, and simple. If your code needs a comment to explain what it does, rewrite the code.

1. When to Use This Skill

• Writing new Python modules, packages, or scripts
• Reviewing Python code for idioms and best practices
• Structuring Python projects for packaging and distribution
• Optimizing Python performance
• Writing async Python code
• Debugging Python-specific issues

2. Modern Python Idioms (3.10+)

Type Hints — Always

# GOOD: Fully typed
def fetch_users(
    team_id: str,
    active_only: bool = True,
    limit: int = 100,
) -> list[User]:
    """Fetch users for a team."""
    ...

# BAD: No types — impossible to maintain
def fetch_users(team_id, active_only=True, limit=100):
    ...

Structural Pattern Matching (3.10+)

# GOOD: Pattern matching for complex dispatch
match command:
    case {"action": "create", "data": data}:
        return create_item(data)
    case {"action": "delete", "id": item_id}:
        return delete_item(item_id)
    case {"action": action}:
        raise ValueError(f"Unknown action: {action}")
    case _:
        raise ValueError("Invalid command format")

Dataclasses & Pydantic Over Raw Dicts

# GOOD: Typed, validated, documented
from dataclasses import dataclass, field
from datetime import datetime

@dataclass
class User:
    id: str
    name: str
    email: str
    role: str = "member"
    created_at: datetime = field(default_factory=datetime.now)

# BETTER for APIs: Pydantic with validation
from pydantic import BaseModel, EmailStr, Field

class UserCreate(BaseModel):
    name: str = Field(min_length=2, max_length=100)
    email: EmailStr
    role: str = Field(default="member", pattern="^(member|admin|viewer)$")

# BAD: Raw dict — no validation, no docs, typo-prone
user = {"name": "Jane", "emial": "[email protected]"}  # typo? who knows