AI agents are only as secure as the skills they can execute.
The OWASP Agentic Skills Top 10 (AST10) focuses on the security risks introduced by agentic skills, the execution layer that can give AI agents real-world capabilities across files, systems, tools, and workflows.…