关于
Frontend Review is an MCP server designed to visually assess UI edit requests. It takes 'before' and 'after' screenshots of a webpage and compares them against a user's edit request description, then provides a 'yes' or 'no' response indicating whether the edit satisfies the request. If the edit fails to meet requirements, the tool offers a detailed explanation to guide further refinement. It supports integration with AI agents like Cursor and Windsurf, and it uses a series of Large Vision Models from Hyperbolic to perform the visual review, with automatic fallback to alternative models if the primary one fails.
主要功能
- Leverages multiple Large Vision Models for robust review processing.
- Supports configuration with various Large Vision Models
- Performs visual reviews of UI edits using screenshots.
- Provides detailed feedback on why edits may not satisfy requests.
- Integrates with AI agents via MCP.
使用案例
- Automated visual validation of frontend changes.
- Ensuring UI edits accurately reflect user requests.
- Integrating visual review into AI-driven development workflows.
Sponsored