AgentSkillsCN

pattern-detection

检测并记录代码库中的现有代码模式。用于在编写新代码时保持与既定惯例的一致性时使用。

SKILL.md
--- frontmatter
name: pattern-detection
description: Detect and document existing code patterns in a codebase. Use to
  maintain consistency with established conventions when writing new code.

Pattern Detection Skill

Purpose

Identify and document patterns for consistency enforcement.

Pattern Categories

1. Naming Conventions

Reference: patterns/naming-conventions.md

Detection method:

  • •Sample 10+ files across different directories
  • •Extract function/class/variable names
  • •Identify dominant patterns

2. Error Handling

Reference: patterns/error-handling.md

Detection method:

  • •Search for try/catch blocks
  • •Find custom error classes
  • •Check for error middleware

3. Testing Patterns

Reference: patterns/testing-patterns.md

Detection method:

  • •Locate test files
  • •Analyze test structure
  • •Identify mocking approach

Consistency Scoring

Rate codebase consistency:

  • •High (90%+): Strong patterns, few deviations
  • •Medium (70-89%): Clear patterns with some variation
  • •Low (<70%): Inconsistent, needs standardization

Output Format

Create patterns document with:

  1. •Detected patterns with examples
  2. •Consistency score per category
  3. •Deviation examples
  4. •Recommendations for standardization

Storage Location

Save to: docs/research/patterns-{date}.md