What problem does it solve? Unprojected Weegloo list API calls return whole documents with every locale bucket of every field, bloating payloads, slowing requests, and overflowing context windows. This Skill teaches how to read Content and Media correctly so queries stay small, fast, and correct. ## Core Features & Use Cases - Projection with select: Apply include/exclude modes, object paths, and order-key rules on every list read, including the agent's own cma_GetList* MCP calls. - Batch and version prefetch: Replace N single GETs with one sys.id[in] list call, and fetch sys.id plus sys.version cheaply before PATCH/PUT optimistic-concurrency updates. - Master/detail and media patterns: Build lightweight list/sidebar UIs with lazy detail fetches, resolve Refer stubs to Media file URLs, and resize images on the fly with /style1../style10 presets. - Search guidance: Decide between in-memory and server-side search, and use the X-Weegloo-Advanced-Search header required for any fields.* filter or order, which MCP tools cannot send. - Use Case: When building a gallery UI, fetch a projected list of ids and titles, lazily load the clicked item's detail, resolve its image Refer to a Media URL, and append /style3 for a 128px WebP thumbnail. ## Quick Start Ask the agent to fetch a list of Weegloo contents projecting only sys.id and fields.title, then load the selected item's detail and resolve its image URL.
