MCP tools reference

Tools are grouped here by what they get done. The prefix on each name tells you what it acts on: document_ for the template library, dmes_ for the document being worked on, and image_, data_, brand_ and conversion_ for images, datasets, brand kits and export.

Rate limit: 30 requests per minute per connected session, across all tools. See Usage policy & limits.

Start from a template

Tool	What it does
`document_SearchTemplates`	Searches brand templates by title keyword and/or category, with facet counts for narrowing
`document_ListTemplateCategories`	Lists categories and their values, for filtering a search
`document_CreateProjectFromTemplate`	Creates a project from a template and returns it, ready to edit

Read a document

The tools most work starts with:

Tool	What it does
`dmes_GetDocumentStructure`	Page count, element counts, and a per-page summary — the right first call on an unfamiliar document
`dmes_GetAllText`	All the text in the document in one call
`dmes_GetAllStyledText`	All text on a page with its formatting, plus block metadata
`dmes_SearchText`	Finds the blocks containing given text
`dmes_GetSmartFields`	Reports which smart fields the document uses and which still need values
`dmes_GetImages`	Every image in the document, with metadata

Lower-level reads

Tool	What it does
`dmes_GetPages`	Page IDs, titles, order, type, element counts
`dmes_GetBlocksOnPage`	Blocks on one page with IDs, bounding boxes, and types
`dmes_GetBlock`	One block: position, size, rotation, type, group, lock state
`dmes_GetTextContent`	Plain text from one block, no formatting
`dmes_GetStyledTextContent`	Formatted text from specific blocks, as styled ranges
`dmes_GetProperty`	A single property of any element — bounding box, fill, rotation, opacity, z-order
`dmes_GetGroupsOnPage`	Groups on a page, with members and combined bounds
`dmes_GetGroupMembers`	A group's member blocks, optionally recursing
`dmes_GetItemsAtPosition`	What sits at an x,y coordinate, in z-order
`dmes_GetItemsInRect`	What sits inside a rectangular region

Change text

Tool	What it does
`dmes_ReplaceStyledText`	Replaces text while preserving formatting — the tool for translation and copy changes
`dmes_ReplaceTextStyles`	Swaps styles document-wide: one font for another, one color for another

Add to a document

Tool	What it does
`dmes_CreatePage`	Adds a page, or duplicates an existing one
`dmes_CreateTextBlock`	Adds a text block with styling
`dmes_CreateImageBlock`	Adds a block with an image already applied
`dmes_CreateShapeBlock`	Adds a vector shape
`dmes_SetBlockImage`	Replaces the image on an existing block

Apply brand and data

Tool	What it does
`brand_GetBrandKit`	Reads the account's colors, fonts and logos by brand role. Its `branding` field passes straight into `dmes_ApplyBranding`
`dmes_ApplyBranding`	Applies a brand kit across every element bound to a brand role
`dmes_ApplyDataSmartFields`	Fills smart fields from dataset rows
`dmes_ApplyUserSmartFields`	Fills user-level smart fields
`dmes_ApplyTeamSmartFields`	Fills team-level smart fields
`dmes_ApplyProjectSmartFields`	Fills project-level smart fields, text and image

Find images

Tool	What it does
`image_SearchImages`	Searches the images the user can reach — owned, shared, and brand — ranked by relevance
`image_GetImage`	Metadata and tags for one image

Work with data

Tool	What it does
`data_ListDataSets`	Datasets the user can reach, with their sources and collections
`data_GetDataSetSchema`	Columns shared across a dataset's collections and columns unique to each, with primary keys
`data_GetCollectionSchema`	Columns, primary keys and item count for one collection
`data_FindDataItems`	Finds rows by primary key or column value, returning refs ready for `dmes_ApplyDataSmartFields`
`data_GetDataSetSamples`	One sample row per collection, for seeing the shape of the data

Export

Tool	What it does
`conversion_ExportDocument`	Renders the document to PDF or an image format and returns a download link

Export takes longer than other calls. Allow at least 60 seconds.