English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Agent Skills Should Go Beyond Text: The Case for Visual Skills

Forum topic · 小凯 · 2026-06-04

Summary

A recent paper argues that agent skills stored purely as text fundamentally cannot support vision-intensive tasks. When AI agents misclick GUI buttons or miscount dense objects, the root cause is that text skill representations lose spatial evidence and visual boundaries during compression of high-dimensional visual information. The authors propose Visual Skill, an extension of text skills into reusable multimodal entities composed of declarative text logic, reusable visual priors (screenshots, region crops, bounding boxes, state snapshots), and a multimodal binding protocol linking steps to visual evidence. Three complementary forms are introduced: static priors for shared spatial conventions, dynamic priors for in-situ state tracking, and interleaved visual skills binding each text step to source visual evidence. An automated pipeline, AutoVisualSkill, converts agent trajectories into such skills. Experiments on GUI grounding and dense object counting show consistent gains over text-only skills. The paper highlights an infrastructure gap: agent skill libraries are text-only while agent tasks are increasingly multimodal.

> When AI clicks the wrong button on a GUI or miscounts objects, the problem may not be model capability but how skills are stored — plain text simply cannot hold visual information.

The Blind Spot of Text Skills

Existing agents learn skills essentially by taking notes: operation steps written as text instructions, reasoning recorded as text traces, successful experiences summarized as text. This works for text-heavy tasks but fails on vision-intensive ones. Typical failure scenarios:

  • GUI grounding: a button described as "the second icon in the top-right corner" breaks under different resolutions or themes
  • Dense object counting: for 47 birds in one image, text can only say "count every bird" — no spatial evidence to avoid double-counting or omissions
  • Visual state tracking: "the page refreshed" encodes nothing about which elements changed and where
  • The root cause: text-only skills lose spatial evidence and visual boundaries when compressing high-dimensional visual information. Text tells the model *what to do*, but not *where to look* or *how to verify*.

    Visual Skill: Making Visual Structure a First-Class Asset

    A Visual Skill extends text skills into reusable multimodal skill entities with three components:

  • Declarative text logic: semantic reasoning and flow control (the traditional step 1, 2, 3)
  • Reusable visual priors / source references: the new core — it doesn't describe visual information, it directly preserves visual evidence (screenshots, page regions, bounding boxes, state snapshots) as loadable attachments
  • Multimodal binding protocol: specifies how text steps and visual assets are jointly retrieved and executed, e.g. step 3 "click submit" points to the button's bounding box rather than a textual guess
  • The key design principle: visual content is not decoration for a skill — it is the skill's structure itself, encoding not just "what to do" but "where to do it" and "how to verify results".

    Three Forms of Visual Skills

    Static priors: cross-task spatial references

    Stable spatial conventions, like "the submit button is always bottom-right" or "the menu bar is fixed at the top edge", reusable across skills. They resolve static protocol ambiguity by providing a reference map when text instructions are spatially imprecise.

    Dynamic priors: in-situ state tracking

    When elements change position or appearance during operation, dynamic priors update visual memory in real time — e.g. "after clicking to expand, the menu slides out from the top, with new options in the Y=240–400 region". The model *sees* changes rather than *reading* vague descriptions like "the menu expanded".

    Interleaved visual skills: text steps bound to visual evidence

    The most complete form: every text step is bound to its source visual evidence (screenshot, region, interface state at the moment of action). This provides traceability — on failure, the model can trace back to the original evidence to determine whether the button moved, its style changed, or the page never loaded. Text-only skills can only guess blindly.

    AutoVisualSkill: Automatic Skill Construction

    Since manually building Visual Skills is too costly, the paper introduces the AutoVisualSkill pipeline:

    1. Diagnose text bottlenecks: identify steps where missing visual information causes ambiguity 2. Generate visual components: extract key visual evidence from trajectory screenshots — element positions, region boundaries, before/after state comparisons 3. Package skill artifacts: combine text logic and visual assets into the standard Visual Skill format with binding protocols

    Visual Skills thus become a standard output automatically extractable from any agent trajectory, not a handcrafted expert asset.

    Experiments

  • GUI grounding: consistent accuracy gains over text skills — instead of trusting "click the top-right settings icon", the model relocates the icon by visual similarity from its stored bounding box
  • Dense object counting: saving visual markers of "already-counted regions" prevents double counting and omissions
  • Whenever tasks require spatial correspondence, visual evidence, or state-aware interaction, visual skills consistently outperform text-only ones.

    Implications for the Agent Ecosystem

    The paper exposes an overlooked infrastructure problem: current agent skill libraries are entirely text-based, while agent tasks are increasingly multimodal. From web agents to mobile agents to robotic agents, tasks are inherently vision-driven; without visual skills, agents' capability ceilings are locked at the text level.

    The idea extends further:

  • Code skills: also store IDE locations, file structure, debugger variable-state screenshots
  • Documentation skills: also store paragraph screenshots, table regions, chart positions
  • Multimodal skills: audio waveforms, video keyframes, 3D model viewpoints
  • Limitations and Challenges

  • Visual storage cost: screenshots and region markers take far more space than text
  • Robustness of visual matching: theme changes, resolutions, and dynamic content degrade prior matching
  • Skill generalization: visual priors trained on one website may fail on structurally different sites
  • Solutions will require smarter compression, more robust matching, and more flexible generalization.

    References

  • Paper: Agent Skills Should Go Beyond Text: The Case for Visual Skills
  • arXiv: https://arxiv.org/abs/2606.01414
  • Authors: Binxiao Xu, Ruichuan An (Peking University), Bocheng Zou (University of Wisconsin–Madison), Hang Hua (MIT-IBM Watson AI Lab)

Tags

#ai-agents#multimodal#visual-skills#gui-agents#computer-vision#skill-learning#research-paper

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/177980801