A
license-
qualityB
maintenanceMCP PoisonGuard protects AI agents from tool description poisoning attacks by scanning tool descriptions and responses in real-time using multiple detection engines, blocking malicious content before it reaches the agent.
Last updated
MIT