lanhu-mcp

Why we included this project

Teams that do most of their work through AI coding assistants but keep their requirements and designs in Lanhu are the ones this server targets. It bridges the two: connect it to Cursor, Claude Code, Windsurf, or Cline and those assistants can read Axure requirement documents and pull design files straight from the collaboration platform. The design parser extracts concrete measurements such as sizes, spacing, colors, and fonts, then returns reference HTML and CSS, so the AI builds against the actual layout instead of guessing. The team knowledge store is what makes it stick: analyzed requirements get cached and shared, so every assistant works from the same context and nobody re-analyzes the same spec on their own. For a small team, that effectively turns Lanhu into the common source of truth for the AI layer.