AI agents select tools from shared registries by matching natural-language descriptions without human verification.

This security gap, documented in a GitHub issue, exposes a major flaw in enterprise agent security.

AI-generated: summaries written by AI from the linked sources. How we use AI