9 matches found
ironcurtain
IronCurtain 一个用于自主AI代理的安全运行时，其安全策略源自人类可读的章程。 当有人写"安全"时，你应该立即持怀疑态度。我们所说的安全是什么意思？ !WARNING 研究原型。 IronCurtain 是一个早期研究项目，探索如何使AI代理足够安全以真正有用。API、配置格式和架构可能会发生变化。欢迎贡献和反馈。 演示 代理被要求克隆一个仓库并推送更改。gitclone 和 gitpush 都被策略引擎升级，但自动审批器自动批准了它们——用户来自命令模式的可信输入（Ctrl-A）提供了明确的意图，因此无需手动 /approve。 问题...
Security Risks in Tool-Enabled AI Agents: A Systematic Analysis of Privileged Execution Environments
Tool-enabled AI agents are increasingly deployed in cloud-hosted environments and offered as services, where they perform side-effecting operations through privileged tools within execution environments. While such agents enable powerful automation, the security implications of hosting autonomous...
CeLLMate: Sandboxing Browser AI Agents
Browser-using agents BUAs are an emerging class of autonomous agents that interact with web browsers in human-like ways, including clicking, scrolling, filling forms, and navigating across pages. While these agents help automate repetitive online tasks, they are vulnerable to prompt injection...
Cross-Site Request Forgery
Cross-Site Request Forgery CSRF is a confused deputy attack where the attacker causes the browser to send a request to a target using the ambient authority of the user’s cookies or network position.1 For example, attacker.example can serve the following HTML to a victim and the browser will send ...
CVE-2020-26975
When a malicious application installed on the user's device broadcast an Intent to Firefox for Android, arbitrary headers could have been specified, leading to attacks such as abusing ambient authority or session fixation. This was resolved by only allowing certain safe-listed headers. Note: This...
SUSE CVE-2020-26975
When a malicious application installed on the user's device broadcast an Intent to Firefox for Android, arbitrary headers could have been specified, leading to attacks such as abusing ambient authority or session fixation. This was resolved by only allowing certain safe-listed headers. Note: This...
CVE-2020-26975
When a malicious application installed on the user's device broadcast an Intent to Firefox for Android, arbitrary headers could have been specified, leading to attacks such as abusing ambient authority or session fixation. This was resolved by only allowing certain safe-listed headers. Note: This...
Session fixation
When a malicious application installed on the user's device broadcast an Intent to Firefox for Android, arbitrary headers could have been specified, leading to attacks such as abusing ambient authority or session fixation. This was resolved by only allowing certain safe-listed headers. Note: This...
CVE-2020-26975
When a malicious application installed on the user's device broadcast an Intent to Firefox for Android, arbitrary headers could have been specified, leading to attacks such as abusing ambient authority or session fixation. This was resolved by only allowing certain safe-listed headers. Note: This...