# Cross-Layer Misalignment Detection in Agent Skills: A Progressive Loading-Aware Contrastive Learning Approach

arXiv: https://arxiv.org/abs/2607.10534 · HTML: https://mlsanity.com/paper/2607.10534/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Agent skill learning and evolution, one of 16 directions in LLM agents and tool use.
- Builds on 20 of its 36 references that are in the map.
- First posted on arXiv 12 Jul 2026.

## Placed in

- LLM agents and tool use / [Agent skill learning and evolution](https://mlsanity.com/agents/agent-skill-learning-and-evolution/)

## Builds on (most cited first, 20 of 20)

- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Supervised Contrastive Learning](https://mlsanity.com/paper/2004.11362/) (2020)
- [BioBERT: a pre-trained biomedical language representation model for biomedical text mining](https://mlsanity.com/paper/1901.08746/) (2019)
- [Don't Stop Pretraining: Adapt Language Models to Domains and Tasks](https://mlsanity.com/paper/2004.10964/) (2020)
- [Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection](https://mlsanity.com/paper/2302.12173/) (2023)
- [SciBERT: A Pretrained Language Model for Scientific Text](https://mlsanity.com/paper/1903.10676/) (2019)
- [WILDS: A Benchmark of in-the-Wild Distribution Shifts](https://mlsanity.com/paper/2012.07421/) (2020)
- [Contrastive Multiview Coding](https://mlsanity.com/paper/1906.05849/) (2019)
- [SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks](https://mlsanity.com/paper/2602.12670/) (2026)
- [Competence-based Curriculum Learning for Neural Machine Translation](https://mlsanity.com/paper/1903.09848/) (2019)
- [AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents](https://mlsanity.com/paper/2406.13352/) (2024)
- [Agent Skills in the Wild: An Empirical Study of Security Vulnerabilities at Scale](https://mlsanity.com/paper/2601.10338/) (2026)
- [Skill-Inject: Measuring Agent Vulnerability to Skill File Attacks](https://mlsanity.com/paper/2602.20156/) (2026)
- [SWE-Skills-Bench: Do Agent Skills Actually Help in Real-World Software Engineering?](https://mlsanity.com/paper/2603.15401/) (2026)
- [Use All The Labels: A Hierarchical Multi-Label Contrastive Learning Framework](https://mlsanity.com/paper/2204.13207/) (2022)
- [EMBER2024 -- A Benchmark Dataset for Holistic Evaluation of Malware Classifiers](https://mlsanity.com/paper/2506.05074/) (2025)
- [SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces](https://mlsanity.com/paper/2605.12015/) (2026)
- [Hierarchical and Contrastive Representation Learning for Knowledge-aware Recommendation](https://mlsanity.com/paper/2304.07506/) (2023)

## Built on by (most cited first, 1 of 1)

- [A Contract-Centered Architecture for Scalable and Manageable Agentic Runtimes](https://mlsanity.com/paper/2608.27086/) (2026)

## Abstract (arXiv)

Large language model (LLM) agents are increasingly extended through Agent Skills, reusable artifacts that package natural-language metadata, procedural instructions, and execution-time resources for runtime use. As open-source skill marketplaces expand, users and agents increasingly rely on brief metadata to select third-party skills, making it difficult to detect inconsistencies between a skill's description and its true behavior, a problem we call cross-layer misalignment. To address this issue, we propose Progressive Loading-Aware Hierarchical Contrastive Learning (PL-HCL), an LLM-based framework that detects misalignment by modeling the layered structure of Agent Skills and learning cross-layer consistency. Using a normalized corpus of over 264,000 open-source skills and a human-verified challenge set, PL-HCL improves Macro-F1 from approximately 0.45 for unadapted baselines to 0.87-0.89 across evaluated LLM backbones. This approach offers an effective screening tool for users and operators, as well as design principles for detecting inconsistencies in layered digital artifacts.