"author:codexguy" matching MCP servers:

Turns a table your coding agent already has — a CSV, dataframe or semantic-model query result — into chart code that fits it first try, by profiling the data locally, ranking roughly 90 chart types by what it can honestly show, and returning verified D3, Plotly or matplotlib source you own and can run anywhere. It can also find the natural views inside a Microsoft Fabric semantic model and say which chart suits each.