What problem does it solve? Charts often fail because they are chosen by habit, built on insufficient data, or shipped without inspection in the final container. This Skill enforces a contract-first workflow so every chart answers a stated question, uses an honest scale, and survives QA at laptop and narrow widths. ## Core Features & Use Cases - Chart Selection Framework: Nine canonical chart families (trend, comparison, composition, distribution, relationship, and more) with a question-to-form mapping table and data sufficiency floors for trends, scatters, and category comparisons. - Chart Contract & Palette Policy: A nine-field contract covering question, takeaway, fields, renderer, and one of three capped color policies, plus encoding rules that ban redundant legends and color-only distinction. - Delivery Surface Routing: Reference guides route output across native chart widgets, HTML fragments, static images, and self-contained HTML reports with static-first fallbacks and progressive runtime upgrades. - Use Case: When asked to turn a quarterly metrics table into a dashboard visual, the Skill writes the contract, checks data sufficiency, picks a ranked bar or dot chart instead of a default line, and QA-checks the rendered result before handoff. ## Quick Start Ask the agent to visualize your dataset, for example: pick the right chart for this weekly signups table and build it for my HTML report.
