What problem does it solve? Analysts using the Heatmap MCP often don't know which of the 13 tools to call, what arguments each accepts, or how to interpret raw JSON-string responses, so analytical questions never become executed queries. ## Core Features & Use Cases - Complete tool reference: Documents all 13 MCP tools with required and optional arguments, output shapes, and the SharedAnalyticsFilterInput filter set. - Seven canonical recipes: Ready-to-run query patterns for page-level RPS ranking, element-level RPC audits, quadrant segmentation, funnel drop-off diagnosis, journey continuation, click-to-revenue heatmap drills, and split-test validation. - Metrics dictionary and gotchas: Maps book terms (RPS, RPC, AOV, CR) to exact API fields and flags pitfalls like unparsed JSON-string fields and derived RPC. - Use Case: Ask "which pages drive the most revenue per session" and get a concrete get_pageview_analytics call with correct metrics, sorting, and interpretation rules. ## Quick Start Ask the assistant to pull your page-by-page revenue per session ranking for the last 28 days using the Heatmap MCP.