AgentSkillsCN

code-analysis

检查代码是否易于非开发者阅读——清晰命名、使用通俗英语注释、无专业术语

SKILL.md
--- frontmatter
name: code-analysis
description: Check if code is readable by non-developers - clear names, plain English comments, no jargon
version: 1.0.0
author: abereyes
triggers:
  - "check readability"
  - "is this code clear"
  - "can non-devs understand"

Code Readability Checker

Analyzes code to ensure non-developers (managers, stakeholders, new team members) can understand it.

What It Checks

  • Clear naming: No cryptic abbreviations (usr_tkn → userToken)
  • Plain comments: Everyday language, not technical jargon
  • Documentation: What/Why/How for major sections
  • Comment ratio: At least 20% of lines should be comments

Usage

bash
python3 analyze.py --path your-file.py --strictness lenient

Example

Bad Code (score: 71/100):

python
def proc(usr, tkn):
    tmp = usr + tkn
    return tmp * 2

Issues: Cryptic names, no comments, unclear purpose.

Good Code (score: 95/100):

python
def process_user_authentication(username, auth_token):
    """Validate user credentials and return auth score"""
    combined_credential = username + auth_token
    return combined_credential * 2

Known Issues

  • May flag false positives in documentation files
  • Works best on actual production code
  • Use --strictness lenient to reduce noise