Back to GitHub trending

QvvvvvvQ/skills_leaks

Collection of extracted skills from popular AI agents product like Kimi, GLM & Minimax

JavaScript25 stars17 forks
View on GitHubShare on X

What it does

The skills_leaks repository provides insights into security vulnerabilities in AI agents, highlighting how users can exploit design flaws to access sensitive skills. Understanding these issues is crucial for developers and researchers to enhance AI security.

Star history

Not enough history yet — 1 day(s) recorded. The daily snapshot builds this up.

Tracking

Last trending
2026-02-21

Creator kit

Hook

Discover how AI agents can unintentionally leak sensitive skills and what it means for security in AI development!

Content angles

  • Create a tutorial on securing AI agents against skill leaks.
  • Discuss the ethical implications of exploiting AI vulnerabilities.
  • Analyze the potential risks of unprotected AI outputs in real-world applications.

Who should care

AI developers, security researchers, and ethical hackers interested in AI vulnerabilities.