Can Agents Trust Their Skills? Uncovering Unsafe Chains of Trust in Skill-Based LLM Agents

LLM agents increasingly rely on installable skills, which are packages of instructions, code, and resources that equip them with task-specific capabilities and, once installed, can be automatically invoked across subsequent user tasks. This creates a chain of trust in which users delegate authori...

EUVD-2023-0105

Malicious code in bioql PyPI...

EPA failure (Access denied)

EPA failure with the following error seen in the logs: 2022-04-13 12:24:19.079 | Tid: 03344 | ERROR | downloadEpaLib | 296 | Failed to verify downloaded EPA library 2022-04-13 12:24:19.079 | Tid: 03344 | DEBUG | nsverifyfile: called 2022-04-13 12:24:19.080 | Tid: 03344 | ERROR | nsverifyTrustedCe...