What specific GUI automation libraries does it support?

While it applies to any GUI automation logic, it contains specific optimization patterns for PyAutoGUI, vision-agent frameworks, and context-optimization tools.

Can it help with vision-based agent accuracy?

Yes, it specifically audits resolution-independent calculations and grid overlay logic to ensure the agent correctly identifies and interacts with UI elements regardless of screen size.

How does this skill differ from a standard AI code review?

Unlike general reviews, this skill focuses on GUI-specific challenges such as coordinate normalization (0.0-1.0 range), screenshot-based debugging, and token-efficient prompt engineering for vision-language models.

When should I trigger the GUI Agent Code Review skill?

Apply this skill when modifying core agent orchestrators, updating system prompts, changing screenshot handling logic, or adding new action types to your automation suite.

GUI Agent Code Review

Name: GUI Agent Code Review
Author: zlyv587

byzlyv587

コラボレーションツール

Reviews GUI automation and vision agent code using domain-specific criteria for prompts, coordinate systems, and context management.

概要

This skill provides specialized code review capabilities for developers building GUI automation and vision-based agents. It moves beyond generic linting to analyze domain-specific challenges such as LLM prompt efficiency, screenshot coordinate normalization, context window management, and PyAutoGUI safety. By integrating automated checklists for debugging and action design, it ensures that vision-based agents are robust, traceable, and token-efficient, helping teams catch subtle logic errors in orchestrators and executors.

主な機能

Evaluate error handling and screenshot logging for enhanced visual debugging
Review coordinate handling to ensure resolution-independent execution across screens
Verify action design safety and PyAutoGUI implementation patterns for reliability
0 GitHub stars
Audit context management and history compression to maintain KV-cache stability
Analyze LLM prompts for token efficiency and structured JSON output consistency

ユースケース

Optimizing LLM system prompts for improved GUI task planning and reliability
Reviewing Pull Requests for complex vision-based automation agents
Debugging coordinate system mismatches in screen-scraping and interaction tools

概要

主な機能

Evaluate error handling and screenshot logging for enhanced visual debugging
Review coordinate handling to ensure resolution-independent execution across screens
Verify action design safety and PyAutoGUI implementation patterns for reliability
0 GitHub stars
Audit context management and history compression to maintain KV-cache stability
Analyze LLM prompts for token efficiency and structured JSON output consistency

ユースケース

Optimizing LLM system prompts for improved GUI task planning and reliability
Reviewing Pull Requests for complex vision-based automation agents
Debugging coordinate system mismatches in screen-scraping and interaction tools