Graphistry MCP

Instructions

Implementation Reference

• src/graphistry_mcp_server/server.py:176-206 (handler)

  The handler function for the 'get_graph_info' tool. It retrieves cached graph data and computes node and edge counts, returning graph metadata. Registered with @mcp.tool() decorator, which also implies schema from signature and docstring.

  @mcp.tool()
  async def get_graph_info(graph_id: str) -> Dict[str, Any]:
      """Get information about a stored graph visualization."""
      try:
          if graph_id not in graph_cache:
              raise ValueError(f"Graph not found: {graph_id}")
  
          graph_data = graph_cache[graph_id]
          edges_df = graph_data["edges_df"]
          source = graph_data["source"]
          destination = graph_data["destination"]
  
          # Get node and edge counts
          if edges_df is not None:
              node_count = len(set(edges_df[source].unique()) | set(edges_df[destination].unique()))
              edge_count = len(edges_df)
          else:
              node_count = 0
              edge_count = 0
  
          return {
              "graph_id": graph_id,
              "title": graph_data["title"],
              "description": graph_data["description"],
              "node_count": node_count,
              "edge_count": edge_count
          }
      except Exception as e:
          logger.error(f"Error in get_graph_info: {e}")
          raise

• src/graphistry_mcp_server/server.py:176-176 (registration)

  Registration of the get_graph_info tool using the FastMCP @mcp.tool() decorator.

  @mcp.tool()