Lanhu MCP Server
An MCP server for the Lanhu (蓝湖) design and requirements platform letting AI coding assistants read design handoff files and requirement documents directly, extract UI assets, translate a design into HTML/CSS reference code, and share notes across a team's AI tools.
Lanhu MCP Server connects AI coding assistants such as Cursor, Windsurf, Claude Code, OpenClaw and Trae to Lanhu, extracting requirement documents and design specifications so an assistant can read them directly instead of a developer copying information across by hand. Requirement analysis extracts a linked Axure page's text and annotations and can present it from a development, testing or quick-exploration point of view, while a long, oversized page is additionally returned as overlapping full-resolution slices with labeled coordinates.
For UI design handoff, it downloads and displays design images, extracts named slices and icon assets, checks their size and hash after download, and returns measurements such as spacing, color and font drawn from the design file alongside a generated HTML and CSS reference, which the project notes still needs checking against the original design and a real render.
A shared team message board lets every connected AI assistant read and write to the same knowledge base, so a requirement analysis or a pitfall found by one developer's AI can be looked up later by a teammate's AI on the same project.