SecureClaw：双边界架构让 LLM 智能体重获控制权

View PDF HTML (experimental)

Abstract:Tool-using large language model (LLM) agents face two distinct security failures: unauthorized external actions and exposure of sensitive plaintext inside the runtime before any final output check can intervene. Existing defenses usually protect one boundary, either the planner/runtime or the point where external actions are executed, and therefore do not by themselves secure both surfaces. We present SecureClaw, a dual-boundary architecture that enforces authorization in a trusted executor and keeps raw protected data outside the LLM runtime. Sensitive reads pass through a trusted gateway that replaces raw values with reference IDs. In the evaluated deployment, the gateway also provides short planning summaries that deliberately release selected information, a process called declassification. Writes that change external state follow a PREVIEW$\rightarrow$COMMIT protocol in which only a trusted executor may execute the exact request authorized by policy. The runtime can still plan over summaries and reference IDs, but cannot directly resolve the IDs into secrets or perform side effects. Across the AgentDojo, AgentLeak, and Agent Security Bench (ASB) benchmarks, SecureClaw is the only defense we evaluate in a common harness that simultaneously retains usable task utility and achieves 0% attack success rate (ASR) on ASB, 0.64% ASR on AgentDojo, and 3.23% overall leak on AgentLeak's attacked parity lane, which measures final-output and internal-relay leakage. Code is available at this https URL.

Submission history

From: Yuhan Ma [view email]
[v1] Mon, 8 Jun 2026 14:29:01 UTC (103 KB)
[v2] Wed, 14 Oct 2026 09:11:06 UTC (186 KB)
[v3] Thu, 15 Oct 2026 09:33:47 UTC (187 KB)