ToolGuardian: Declarative Security for AI Agent-Tool Interactions

LLM agents increasingly rely on external tools, expanding capability while creating a new security boundary: third-party tools may appear benign at the interface level while embedding unsafe behavior in implementation. Existing defenses rely on weak metadata, collapse characterization and policy...

MAL-2025-112256 Malicious code in respective_salamander_amethyst-57 (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 4732193a365b795a76d14c67b280a999744de63cc15da083f3d9e2dc98972c77 This package appears to be part of the tea.xyz token reward campaign that flooded npm. These packages typically contain autopublish scripts auto.js,...

Malicious code in greasy_bat_z3n (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 116bf5bf3ebb36c5d36e84ee6399c59a55e35b927229c84e61ca8d65608e2f05 This package appears to be part of the tea.xyz token reward campaign that flooded npm. These packages typically contain autopublish scripts auto.js,...

Warning — 5 New Trojanized Android Apps Spying On Users In Pakistan

Cybersecurity researchers took the wraps off a new spyware operation targeting users in Pakistan that leverages trojanized versions of legitimate Android apps to carry out covert surveillance and espionage. Designed to masquerade apps such as the Pakistan Citizen Portal, a Muslim prayer-clock app...