tool poisoning
AI security interview questions tagged tool poisoning, across every topic.
3 questions · 1 unlocked for you
Concepts behind "tool poisoning"
Explanations of the mechanisms these questions rely on.
Foundational
MCP SecurityMCP connects an AI application to servers offering tools, resources and prompts. Review what the host puts into model context, what local processes can reach, and which identities and permissions govern remote calls. Metadata approval, transport authentication and authorization of a concrete operation are separate controls. Agent & MCP Security
Core
Tool Poisoning and Rug PullsA server can place unwanted instructions in the tool metadata a host exposes to its model. Changed definitions can invalidate an earlier review, and one tool description can influence calls to another server. Review definition revisions before use, confine implementations and authorize each consequential action independently; a matching metadata hash does not prove unchanged behavior. Agent & MCP SecuritySign in