No user-submitted related servers found.
Scored across 94 tools
The set contains many near-duplicate families—wb_move vs wb_live_move, wb_scale vs wb_live_scale, wb_duplicate vs wb_live_duplicate, and several scatter/array variants—so an agent must carefully distinguish saved-build edits from live-editor actions. The build/native/worker import pipeline also has multiple overlapping entries such as wb_native_import, wb_build_native_import, wb_worker_import, and wb_build_worker_import. Descriptions help, but tool boundaries are still easy to misselect.
All tools are snake_case with a consistent wb_ prefix and readable subsystem names such as live, build, native, worker, and assets. However, verb/noun ordering is inconsistent: wb_assets_search and wb_build_prepare versus wb_spawn_asset and wb_plan_apply, plus bare noun-style names like wb_tree and wb_refs. The set is readable but not fully predictable.
At 94 tools this is far beyond a well-scoped MCP surface, even for a broad World Builder automation domain. The set is fragmented into highly granular live/offline and pipeline variants that could be consolidated or grouped behind fewer tools.
The surface covers an unusually full lifecycle: asset cataloging, semantic search, bounds and collision checks, planning, generation, live editing, export inspection, native conversion, packaging, verification, and deployment. Remaining gaps are minor—for example, offline saved-build deletion/rename and favorite removal are not clearly exposed—but core workflows are well covered.