GitHub Stars is the Worst Metric for Skill Quality
GitHub Stars is the Worst Metric for Skill Quality
This is an English translation of the original Chinese article.
When evaluating AI skills, most people look at GitHub Stars. This is a mistake.
The Problem with Stars
GitHub Stars measure popularity, not quality. Here's why this matters:
- Viral effects: A skill shared on Twitter gets stars regardless of quality
- Age bias: Older projects accumulate more stars over time
- Marketing matters: Well-marketed skills outshine better alternatives
- No usage correlation: Stars don't mean people actually use it
What You Should Look At Instead
- Active development: Recent commits and updates
- Issue resolution: How quickly bugs are fixed
- Documentation quality: Clear README and examples
- Real-world usage: Actual installations and calls
- Community feedback: Reviews and recommendations
The skillsAgent Rating System
We built a better rating system that considers:
- Code quality and maintainability
- Documentation completeness
- Real usage statistics
- User reviews and ratings
Stop counting stars. Start measuring what matters.