Skill Claude CodeCodex
End-to-end Simba MMM workflow over MCP — upload a dataset, create and poll a Bayesian model, and read results correctly (section semantics, channel naming, attribution/Overlap rules, context-size controls). Use when building or analyzing an MMM through the Simba MCP tools.
Skill Claude CodeCodex
Run Simba budget optimizations correctly over MCP — payload conventions (percent bounds, laydown/CPM arrays), revenue vs profit objectives, polling by runid, interpreting decision vs comparison columns, and curating run history. Use when optimizing budgets or reading optimizer results through the Simba MCP tools.
Skill Claude CodeCodex
Simba prior-override payload conventions for createmodel — smart-default merging, strict field rejection, the half-saturation/half-marginal/half-life anchor families and which combinations are invalid. Use before constructing any priors[] override through the Simba MCP tools.
Skill Claude CodeCodex
Long-term (VAR) modeling workflow over Simba MCP — create a VAR model, poll its longer fit, link it to an MMM, and read the combined long-run rollup. Use when quantifying long-term/brand-equity effects beyond the MMM's short-term window.
MCP server Claude CodeCodexCursor +2
MCP server for Simba Marketing Mix Modeling — connect AI assistants to your MMM models. Runs locally from the simba-mcp Python package.