# ContractSkill: Repairable Contract-Based Skills for Multimodal Web Agents

arXiv: https://arxiv.org/abs/2603.20340 · HTML: https://mlsanity.com/paper/2603.20340/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use.
- Builds on 27 of its 41 references that are in the map.
- First posted on arXiv 20 Mar 2026.

## Placed in

- LLM agents and tool use / [Web, search, and research agents](https://mlsanity.com/agents/web-search-and-research-agents/)

## Builds on (most cited first, 27 of 27)

- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [Mind2Web: Towards a Generalist Agent for the Web](https://mlsanity.com/paper/2306.06070/) (2023)
- [OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments](https://mlsanity.com/paper/2404.07972/) (2024)
- [GPT-4V(ision) is a Generalist Web Agent, if Grounded](https://mlsanity.com/paper/2401.01614/) (2024)
- [AppAgent: Multimodal Agents as Smartphone Users](https://mlsanity.com/paper/2312.13771/) (2023)
- [WorkArena: How Capable Are Web Agents at Solving Common Knowledge Work Tasks?](https://mlsanity.com/paper/2403.07718/) (2024)
- [SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks](https://mlsanity.com/paper/2602.12670/) (2026)
- [Reinforcement Learning on Web Interfaces Using Workflow-Guided Exploration](https://mlsanity.com/paper/1802.08802/) (2018)
- [SkillWeaver: Web Agents can Self-Improve by Discovering and Honing Skills](https://mlsanity.com/paper/2504.07079/) (2025)
- [Mobile-Agent-v2: Mobile Device Operation Assistant with Effective Navigation via Multi-Agent Collaboration](https://mlsanity.com/paper/2406.01014/) (2024)
- [Agent S: An Open Agentic Framework that Uses Computers Like a Human](https://mlsanity.com/paper/2410.08164/) (2024)
- [WebLINX: Real-World Website Navigation with Multi-Turn Dialogue](https://mlsanity.com/paper/2402.05930/) (2024)
- [The BrowserGym Ecosystem for Web Agent Research](https://mlsanity.com/paper/2412.05467/) (2024)
- [AutoWebGLM: A Large Language Model-based Web Navigating Agent](https://mlsanity.com/paper/2404.03648/) (2024)
- [AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents](https://mlsanity.com/paper/2410.13825/) (2024)
- [Inducing Programmatic Skills for Agentic Tasks](https://mlsanity.com/paper/2504.06821/) (2025)
- [WorkArena++: Towards Compositional Planning and Reasoning-based Common Knowledge Work Tasks](https://mlsanity.com/paper/2407.05291/) (2024)
- [AutoManual: Constructing Instruction Manuals by LLM Agents via Interactive Environmental Learning](https://mlsanity.com/paper/2405.16247/) (2024)
- [Group-Evolving Agents: Open-Ended Self-Improvement via Experience Sharing](https://mlsanity.com/paper/2602.04837/) (2026)
- [When Single-Agent with Skills Replace Multi-Agent Systems and When They Fail](https://mlsanity.com/paper/2601.04748/) (2026)
- [InSTA: Towards Internet-Scale Training For Agents](https://mlsanity.com/paper/2502.06776/) (2025)
- [PolySkill: Learning Generalizable Skills Through Polymorphic Abstraction](https://mlsanity.com/paper/2510.15863/) (2025)
- [Retrieval-augmented GUI Agents with Generative Guidelines](https://mlsanity.com/paper/2509.24183/) (2025)
- [PG-Agent: An Agent Powered by Page Graph](https://mlsanity.com/paper/2509.03536/) (2025)
- [Enhancing Web Agents with a Hierarchical Memory Tree](https://mlsanity.com/paper/2603.07024/) (2026)
- [ParamMem: Augmenting Language Agents with Parametric Reflective Memory](https://mlsanity.com/paper/2602.23320/) (2026)

## Built on by (most cited first, 13 of 13)

- [SkillGenBench: Benchmarking Skill Generation Pipelines for LLM Agents](https://mlsanity.com/paper/2605.18693/) (2026)
- [Dynamic Agent Skills: A Lifecycle Survey and Taxonomy of Evolving Skill Libraries](https://mlsanity.com/paper/2607.10113/) (2026)
- [Online Skill Learning for Web Agents via State-Grounded Dynamic Retrieval](https://mlsanity.com/paper/2606.04391/) (2026)
- [Ace-Skill: Bootstrapping Multimodal Agents with Prioritized and Clustered Evolution](https://mlsanity.com/paper/2605.08887/) (2026)
- [Skim: Speculative Execution for Fast and Efficient Web Agents](https://mlsanity.com/paper/2605.16565/) (2026)
- [Harnessing Agent Skills: Architectural Patterns and a Reference Architecture for Skill-Mediated LLM Agents](https://mlsanity.com/paper/2606.20631/) (2026)
- [EconSkills: Studying Skill Transfer and Retrieval for Web Agents on Live Economic Data](https://mlsanity.com/paper/2609.19523/) (2026)
- [Beyond Domains: Reusing Web Skills via Transferable Interaction Patterns](https://mlsanity.com/paper/2606.17645/) (2026)
- [Attributing Structured-Output Gains in Function Calling: Interface Alignment versus Procedural Transfer](https://mlsanity.com/paper/2607.02595/) (2026)
- [Repo2Skill-Evo: Repository Skills Go Stale in Silence](https://mlsanity.com/paper/2608.21964/) (2026)
- [Towards a Systems Foundation for Agentic Skills: Architecture, Lifecycle, and Security](https://mlsanity.com/paper/2608.29596/) (2026)
- [SkillAlign: Aligning Skill Interfaces for LLM-based Agents](https://mlsanity.com/paper/2609.07255/) (2026)
- [Trajectory-Guided Fault Localization for Agent Skill Evolution](https://mlsanity.com/paper/2610.11858/) (2026)

## Abstract (arXiv)

Self-generated skills for web agents are often unstable and can even hurt performance relative to direct acting. We argue that the key bottleneck is not only skill generation quality, but the fact that web skills remain implicit and therefore cannot be checked or locally repaired. To address this, we present ContractSkill, a framework that converts a draft skill into an executable artifact with explicit procedural structure, enabling deterministic verifica tion, fault localization, and minimal local repair. This turns skill refinement from full rewriting into localized editing of a single skill artifact. Experiments on VisualWebArena show that Contract Skill is effective in realistic web environments, while MiniWoB provides a controlled test of the mechanism behind the gain. Under matched transfer layers, repaired artifacts also remain reusable after removing the source model from the loop, providing evi dence of portability within the same benchmark family rather than full-benchmark generalization. These results suggest that the central challenge is not merely generating skills, but mak ing them explicit, executable, and repairable. Code is available at https://github.com/underfitting-lu/contractskill.git.