What problem does it solve? When an agent produces analytical results — metrics breakdowns, billing investigations, security audits, or structured MCP tool data — plain markdown tables and code blocks in chat are hard to read and easy to lose. This Skill renders that output as a live React canvas the user can open beside the chat in Cursor IDE. ## Core Features & Use Cases - Canvas decision workflow: Determines when output warrants a standalone canvas artifact versus plain chat text, based on user intent rather than response shape. - Cursor canvas SDK guidance: Enforces correct use of the cursor/canvas component library — charts, tables, cards, diffs, forms, DAG layouts — with type declarations for every export. - Design quality enforcement: Applies a flat, minimal design policy (no gradients, emojis, box-shadows, or empty states) with a pre-delivery self-check. - Use Case: After querying Datadog via MCP for API error rates, render the results as a labeled multi-series line chart canvas with stats and a findings table instead of dumping a markdown table into chat. ## Quick Start Ask the agent to analyze the last 7 days of Datadog error rates and present the results as a canvas.
