This one builds the kind of comp tables you'd see in an investment banking pitch book or IC memo. It's opinionated about structure: formulas over hardcoded values, explicit data source hierarchy (MCP sources like S&P Kensho first, never web search for financials), and step-by-step verification so you catch errors before the whole sheet is done. The approach is practical: ask about audience and context first, then adapt the metrics and layout accordingly rather than stamping out identical templates. Best for public company valuation work where you need operating metrics, multiples, and statistical benchmarking that update cleanly when inputs change. Not suitable for pre-revenue startups or companies without clean public comps.
npx skills add https://github.com/anthropics/financial-services-plugins --skill comps-analysis