Python

Write reliable Python avoiding mutable defaults, import traps, and common runtime surprises.

Security Scan

This is an instruction-only Python style/best-practices helper (no code or installs) that only requires python3 and otherwise behaves consistently with its stated purpose.

Purpose & Capability

Name, description, and required binary (python3) align with a Python guidance helper. The content is documentation on common Python pitfalls and does not request unrelated credentials or tools. Minor metadata inconsistency: registry slug is 'py' while SKILL.md uses slug 'python' — likely cosmetic and not a security issue.

Instruction Scope

SKILL.md and the included .md files are documentation-only; they do not instruct the agent to read arbitrary files, send data externally, or run network requests. There are no open-ended directives that would give the agent broad discretion to collect or exfiltrate data.

Install Mechanism

No install spec and no code files to write or execute. This is the lowest-risk form (instruction-only).

Credentials

The skill requests no environment variables, credentials, or config paths. Requesting only python3 is proportionate for a Python guidance skill.

Persistence & Privilege

Defaults (always: false, model invocation allowed) are appropriate. The skill does not request permanent presence or system-wide changes.

Assessment

This skill is documentation-only and appears coherent with its purpose. It doesn't include executable code or ask for secrets. Before installing, confirm your agent sandbox: if the agent is permitted to execute python3 or access local files, the agent could (independently of this skill) run Python code you provide — that's a general platform permission concern, not an issue with this skill. If you want extra caution, keep the agent's execution rights and file access restricted, and verify the system python3 version matches what you expect.

Quick Reference

Topic	File
Dynamic typing, type hints, duck typing	types.md
List/dict/set gotchas, comprehensions	collections.md
Args/kwargs, closures, decorators, generators	functions.md
Inheritance, descriptors, metaclasses	classes.md
GIL, threading, asyncio, multiprocessing	concurrency.md
Circular imports, packages, init.py	imports.md
Pytest, mocking, fixtures	testing.md

Critical Rules

• def f(items=[]) shares list across all calls — use items=None then items = items or []
• is checks identity, == checks equality — "a" * 100 is "a" * 100 may be False
• Modifying list while iterating skips elements — iterate over copy: for x in list(items):
• GIL prevents true parallel Python threads — use multiprocessing for CPU-bound
• Bare except: catches SystemExit and KeyboardInterrupt — use except Exception:
• UnboundLocalError when assigning to outer scope variable — use nonlocal or global
• open() without context manager leaks handles — always use with open():
• Circular imports fail silently or partially — import inside function to break cycle
• 0.1 + 0.2 != 0.3 — floating point, use decimal.Decimal for money
• Generator exhausted after one iteration — can't reuse, recreate or use itertools.tee
• Class attributes with mutables shared across instances — define in __init__ instead
• __init__ is not constructor — __new__ creates instance, __init__ initializes
• Default encoding is platform-dependent — always specify encoding='utf-8'