Hardware Keystores for AI Agent Signing Workflows: A Zero-Trust MCP Enforcement Architecture

• 类型：arxiv
• 标识：2608.06130
• 链接：http://arxiv.org/abs/2608.06130v1
• 主分类：agent
• 形态：application
• 被引：0
• 被引来源：Semantic Scholar
• S2被引：0
• 影响力被引：0
• TLDR：This work replaces software-resident keys with hardware-confined keys accessible through a vendor-neutral PKCS#11 interface, enabled by a surrounding five-layer Zero-Trust enforcement stack comprising session identity, scope bounds, semantic validation, taint tracking, and the hardware execution boundary.
• 待LLM分类：否
• 标题中文：面向 AI Agent 签名工作流的硬件密钥存储：一种零信任 MCP 强制执行架构
• TLDR中文：将软件侧密钥替换为硬件受限的密钥，通过厂商中立的 PKCS#11 接口访问，并由包含会话身份、作用域边界、语义验证、污点追踪和硬件执行边界的五层 Zero-Trust 执行栈提供保护。
• 来源文件：
• /inbox/tom/_candidates/2026-08-07-agent-rag-longcontext-candidates.json
• /inbox/tom/_candidates/2026-08-08-agent-rag-longcontext-candidates.json
• [S2 enrich]