This gives Claude visibility into your system architecture and lets it participate in engineering decisions. The focus is on helping AI agents understand how your components fit together and maintain architectural constraints as they work. You'd reach for this when you want Claude to grasp the bigger picture of your codebase structure, suggest changes that respect your architectural boundaries, or help document relationships between services. It's built for teams where the agent needs to understand not just individual files but how the whole system is organized. The stdio transport means it runs locally alongside your development environment.
claude mcp add --transport stdio io.gjalla-mcp-server -- uvx gjalla