# Paper: SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents

---
type: paper
title: "SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents"
authors: Yuchuan Tian, Mengyu Zheng, Haocheng Mei, Ye Yuan, Chao Xu, Xinghao Chen, Hanting Chen, Yu Wang
year: 2026
venue: arXiv
url: https://arxiv.org/abs/2606.18356
code_url:
source: arxiv
collected_at: 2026-07-08
published_at: 2026-06-16
updated_at: 2026-06-16
status: queued
relevance: high
topics:
  - agent-evaluation
  - agent-safety
  - computer-use
  - memory
  - tool-use
methods:
  -
benchmarks:
  -
models:
  -
datasets:
  - cs.CR
  - cs.AI
related_concepts:
  -
related_jobs:
  -
related_experiments:
  -
related_projects:
  -
collection_score: 19
collection_queries: agent-safety, tool-use
---

## One-line Takeaway

Auto-collected from arXiv because it matched the Agent collection queries. Needs human skim.

## Why Collected

- matched queries: agent-safety, tool-use
- inferred topics: agent-evaluation, agent-safety, computer-use, memory, tool-use
- arXiv categories: cs.CR, cs.AI
- collection score: 19

## Review Checklist

- Does this paper directly inform Agent architecture, evaluation, memory, tools, safety, coding agents, GUI/browser agents, or multi-agent workflows?
- Does it include a benchmark, dataset, code, or reproducible experimental setup?
- Should it be promoted from `queued` to `skimmed` or `summarized`?

## Links

- arXiv: https://arxiv.org/abs/2606.18356