Daily desk

News

Artificial intelligence, technology, science, health and education β€” reported daily, with the original source attached to every item.

1 story

A donut chart shows 17.6% of the 954 audited AI agent skills flagged for exploitable prompt-injection flaws, with a cracked skill-file icon above.
Latest πŸ€– Artificial Intelligence

Study Finds 17.6% of Nearly 1,000 Real AI Agent Skills Carry Exploitable Prompt-Injection Flaws

A new arXiv study audited 954 AI agent skills from the marketplace skills.sh and found 168, or 17.6%, carried exploitable prompt-injection flaws β€” one of the first systematic security audits of a fast-growing, largely unreviewed distribution model.

EduFabTech Β· 16h ago Β· 4 min read