Unverified50% confidenceConceptExact time
Script security is divided into two situations: the risk of scripts auto-generated by the agent at runtime, and the risk introduced by third-party plugins and Skills.
1
Sources
50%
Confidence
Long-term
Relevance
9/28/2026
First Seen
Sources
Related Entities
Related Claims
Unverified无载荷技能攻击通过技能的元数据、描述文本、命名结构或调用逻辑的组合，诱导智能体在特定上下文中触发非预期的危险行为68% similarUnverifiedBuilding enhancement tools around leading platforms carries platform risk: if the underlying platform adjusts its policies or interfaces, the functionality of third-party tools may be directly affected.68% similarUnverifiedFor security reasons, Skills scripts should be executed in isolation, running in a controlled sandbox environment rather than in the agent's main process.67% similarUnverified技能可以包含并运行代码，下载陌生人的技能等同于在自己机器上运行陌生人的代码，构成安全风险66% similarUnverifiedAn agent's performance in real-world security operations depends on the accuracy of its decisions, the richness of its tool-calling capabilities, and its ability to adapt to complex environments.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/957362API
curl https://kongchang.com/api/v1/knowledge/claims/957362MCP
get_claim(id=957362)