ContractSkill: Repairable Contract-Based Skills for Multimodal Web Agents

Abstract

Despite rapid progress in multimodal GUI agents, reusable skill acquisition remains difficult because on-demand generated skills often leave action semantics, state assumptions, and success criteria implicit. This makes them brittle to execution errors, hard to verify, and difficult to repair. We present ContractSkill, a framework that converts a draft skill into a contracted executable artifact with explicit preconditions, step specifications, postconditions, recovery rules, and termination checks. This representation enables deterministic verification, step-level fault localization, and minimal patch-based repair, turning skill refinement into localized editing rather than full regeneration. Experiments on VisualWebArena and MiniWoB with GLM-4.6V and Qwen3.5-Plus show that ContractSkill improves self-generated skills from 9.4% and 10.9% to 28.1% and 37.5% on VisualWebArena, and from 66.5% and 60.5% to 77.5% and 81.0% on MiniWoB. Repaired artifacts also transfer across models, improving the target model's self-generated-skill baseline by up to 47.8 points and 12.8 points on the two benchmarks, respectively. These results suggest that agent skills are better treated as explicit procedural artifacts that can be verified, repaired, and shared across models.

Figures (4)

• Figure 1: Overview of ContractSkill. The pipeline proceeds through five stages. It starts with input and draft generation, then compiles the draft into a contracted artifact, performs verification-guided execution with deterministic fault localization, applies minimal patch repair under an explicit objective, and finally reuses the repaired artifact with both the source model and a target model.
• Figure 2: Representative repair case on VisualWebArena. The verifier localizes the failure to the post-exploration stage, and ContractSkill repairs the artifact by inserting only the missing continuation steps rather than rewriting the whole strategy.
• Figure 3: Template-level MiniWoB pass rates across the main baselines and ablations. ContractSkill gains concentrate on a small subset of repairable templates.
• Figure 4: Supplementary VWA failure analysis across Qwen3.5-Plus and GLM-4.6V. The top panel shows verifier error codes among failed episodes. The bottom panel shows the first failing step bucket. Across both models, ContractSkill reduces the dominance of late coarse failures that are common in No-Skill execution and exposes a larger share of residual failures as earlier, more localizable breakdowns.