Security researchers analyzed 10,000 AI skills and discovered widespread vulnerabilities to code exploits and prompt injection attacks.

SafeSkill conducted a large-scale security audit examining 10,000 AI skills for common attack vectors

The study identified significant vulnerabilities related to code exploits and prompt injection techniques

Findings highlight critical security gaps in AI skill development and deployment

Research aims to raise awareness about AI safety and secure coding practices in skill development

Results published on SafeSkill.dev platform for community reference and improvement