Stack layer / Contrast
"Not Worth Another Token": Pruning Deep-Research Agent Context Before Retrieval Cuts Token Usage Up to 73% With Little Quality Loss
arXiv (via HuggingFace Daily Papers)
Stack layer / Threat pattern
12,110 .cursorrules Files From 11,427 Repos: Adoption Concentrates in Single-Maintainer Toy Projects, and Security Guidance Is Rare
arXiv 2608.10622
Policy dependency / Stack layer
Split your agent's safety into four evolvable artifacts — system prompt, rule bank, safety memory, tool policy — for a 3.1x attack-success reduction
arXiv 2608.09885
Policy dependency / Stack layer
RangeBench: 1,148 Multi-Hop Cyber Ranges Show Attack Agents Get a Foothold Then Stall — 24.5–47.0% Never Finish the Chain
arXiv 2608.09526
Stack layer / Contrast
Programmable VLM Backdoor: One Poisoning Phase Lets an Attacker Pick Unseen Caption Targets at Inference Time
arXiv 2608.10959
Policy dependency / Stack layer
POLIS 5,280-Episode Study: Provenance-Aware Guards Block Authority Laundering That Local-State Guards Miss in 22 of 96 Episodes
arXiv 2608.09828
Stack layer / Contrast
Deterministic rule-guided dispatch beats autonomous code-review agents by 2.17x SEM-F1 on 5–15x fewer tokens
arXiv 2608.09290
Stack layer / Contrast
Top-Down Differential Analysis Finds GCC and Clang Produce Substantially Different Performance From Identical Source — Then Patches Binaries by Transplanting the Winner's Code
arXiv 2608.09530