EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Aug 25·5 min read
What languages are agent skills written in?
TL;DR
In early 2026, 14.3% of newly created AI agent skills were written in languages other than English, indicating a significant shift in global development. This trend highlights the rapid growth of non-English contributions, particularly from countries like China and India.
✦ Why It Matters
Consider developing AI skills in multiple languages to reach a broader audience and enhance usability.
Key Takeaways
- Monitor the rise of non-English skills in your projects.
- Explore collaboration with developers from diverse linguistic backgrounds.
- Adapt documentation practices to accommodate multilingual contributions.
- Stay informed about global trends in AI development.
Full Summary
Agent skills, defined as SKILL.md files containing instructions for AI agents, have seen a notable increase in non-English contributions on GitHub. In the first quarter of 2026, 13.0% of newly written skills were in languages other than English, rising to 16.3% by the second quarter.
This contrasts sharply with the slow growth of non-English documentation on GitHub over the past decade. The analysis revealed that while English remains dominant at 85.3%, languages like Chinese (6.2%) and various European languages are gaining ground.
The data was collected from 3.8 million skills across 282,200 public repositories, highlighting a shift in the global landscape of AI development, particularly as countries like India and Brazil contribute significantly to the developer population. This trend indicates that the AI ecosystem is becoming increasingly multilingual and diverse.