No user-submitted related servers found.
Scored across 20 tools
Most tools map to a distinct resource plus action (list_decks, get_deck, delete_deck, add_asset, list_assets), and the descriptions draw sharp lines — e.g. apply_ops is explicitly the only mutation path and get_deck_view is flagged as never needing a call. The only friction is around inspection: render_preview, validate_deck, get_deck, get_deck_view and create_deck's returned findings all touch 'checking a deck', so an agent must read descriptions carefully to pick.
Every tool is snake_case verb_noun: create_deck, list_decks, get_deck, apply_ops, render_preview, export_deck, revert_deck, delete_deck, create_brand_kit, list_brand_kits, get_brand_kit. The noun for each resource (deck, asset, brand_kit, example) is used consistently across verbs, so an agent can predict unprefixed names.
20 tools is at the upper edge but each functional area earns its slot: deck lifecycle, brand kits, assets, validation/preview, export, history, examples. The one weak entry is get_deck_view, an internal chat-render helper that adds surface without adding capability.
The deck lifecycle is essentially complete (create, list, get, apply_ops for all mutations, history, revert, delete, plus validate/render/export). However, secondary resources are half-covered: brand kits can be created, listed and fetched but never updated or deleted, and assets can be added and listed but not fetched individually or removed.