What problem does it solve? Planning work on an Umbraco 17 project is error-prone when done blind: schema IDs, existing compositions, and backoffice extension registry patterns cannot be reliably inferred from training data or committed files alone. This Skill supplies stack-specific planning rules so plans are grounded in the live CMS state and authoritative extension guidance. ## Core Features & Use Cases - Live schema inspection before planning: Directs use of Umbraco MCP tools (get-all-document-types, get-all-data-types, and related lookups) to surface real IDs, property editors, and reusable compositions before designing schema steps. - Backoffice extension routing: Maps each extension type (dashboards, property editors, workspaces, modals, trees, context API, entry points, entity actions, block editor custom views) to the authoritative umbraco-cms-backoffice-skills companion, with explicit fallback behavior when those skills are unavailable. - Layer vocabulary and step ordering: Defines the layers an Umbraco feature spans (schema, slice, client-side components, backoffice extension, tests) and the step order that usually works, including model regeneration handling. - Use Case: When asked to plan a new document type with a custom dashboard, first query the live instance for existing compositions, record discovered IDs in Key Decisions, then sequence schema, slice, and extension steps per the prescribed order. ## Quick Start Ask the assistant to plan an Umbraco 17 feature such as adding a document type with a custom backoffice dashboard, and it will inspect the live schema and sequence the work steps accordingly.
