Available tools

The tools your assistant uses to read and edit your board, grouped by what they do.

Your assistant chooses which tools to use when you describe a task. One request may need several tools: reading the board, making changes, then checking the result.

This page lists every tool by the name shown in AI-app permission prompts and in Agent activity. The local board MCP server and Board WebMCP offer the same tools, although file access and some defaults differ.

Read and review: inspect boards, shapes, documents, image layers, and layout problems.
Create and edit: add content, change its appearance, and organize shapes or media.
Animate and configure: build animations, set up cameras and lenses, adjust settings, and export images.
Find help: look up documentation and icons, or use a script for a task that needs custom logic.

Reading the board

These tools inspect content without editing it. Reading the live editor state or taking a board screenshot pauses animation playback.

Tool	What it does
`list_boards`	Lists the boards saved in this browser and the actions available on each one, with filters and paged results.
`read_board`	Reads a board's pages, shapes, named selections, saved settings, or the files stored on it.
`read_session`	Reads the open board and active Version, whether editing is allowed, your interface preferences, the current selection, any selected image area, and the area in view.
`read_agent_context`	Gets a starting overview in one step: the tab's state, a summary of the selected shapes, and an optional picture of the current view.
`view_board`	Takes a picture of part of the board and lists the visible shapes and their bounds, with optional animation path or pivot overlays.
`inspect_shape`	Reads one shape's content, appearance, connected arrows, and animation blocks. It can also return an image of the shape, its original file, or its attachments.
`review_board_area`	Checks an area of the board for layout problems, such as overlapping text, cut-off labels, low contrast, or disconnected arrows.
`read_document_content`	Reads text from PDFs and e-books, searches text, and reads PDF annotations. It also returns page images or sheets of thumbnails from PDFs, e-books, and comics, including documents in table cells.
`read_result_page`	Reads the next part of a long result without repeating the original operation.

For web embeds, online videos, or HTML content, the assistant may need you to share live content before it can capture what is displayed.

Managing boards

This tool works with the boards saved in this browser.

Tool	What it does
`manage_boards`	Creates, opens, renames, duplicates, reorders, or deletes boards saved in this browser.

Deleting a board: asks you to confirm in the Irakasi tab before permanently removing its data from this browser.
Outside board undo: managing boards is outside the board's undo history.

Creating content

These tools add new shapes and files to the board.

Tool	What it does
`create_shapes`	Creates editable shapes, including geometry, text, rich text, text on a path, tables, code, frames, cameras, vector drawings, arrows, web embeds, and HTML shapes. It also places imported images, videos, or file cards, and creates images from SVG code.
`import_media`	Imports images, video, audio, PDFs, e-books, comics, and other files for use on the board or in table cells. It can place an image, audio card, or book directly on the board.

The assistant can create up to 200 shapes at once in one undo step, or 100 when it supplies SVG code. Image generation depends on your AI app; import_media brings the resulting image into Irakasi.

Editing shapes

These tools change existing shapes on the board.

Tool	What it does
`edit_shape_transform_and_state`	Moves, nudges, resizes, rotates, or flips shapes, changes their pivot, aspect-ratio lock, proportional stroke scaling, lock, or visibility, or duplicates them.
`edit_shape_appearance`	Changes opacity, outlines, fills, corners, gradients, image fills, and shape filters. It also changes the appearance of geometry, frames, cameras, and supported document attributes in table cells.
`edit_shape_text_and_content`	Edits text and typography, table content and structure, media playback, text-on-path placement and spacing, document view and reading position, web embeds, HTML, or code.
`edit_shape_media`	Shares or replaces media, crops images and videos, edits table-cell media, or extracts document pages as images. It also converts SVG images to editable shapes and converts between media and file-card formats.
`edit_shape_relationships`	Groups or ungroups shapes, moves them into or out of frames or cameras, fits frames to their content, and sets up or edits shape masks.
`replace_vector_geometry`	Replaces the vector outline or points of a path, freehand, or polyline shape.
`edit_arrow_shape`	Changes an arrow's style, routing, handles, labels, or connections to other shapes.
`arrange_shapes`	Aligns shapes, arranges them in rows, columns, or grids, stretches them, or changes their front-to-back stacking order.
`boolean_shapes`	Combines vector shapes or separates them into parts: union, subtract, intersect, exclude, flatten, or split.
`convert_shapes`	Converts shapes into paths, freehand paths, arrows, rich text, or images.
`edit_named_selections`	Lists, saves, selects, renames, updates, or deletes named selections. It can add or remove members and reorder both the selections and the shapes within them.
`delete_board_entities`	Deletes shapes, animation blocks, or pauses.

The order of shapes in a named selection also sets the stagger order for animation blocks that use it.

Images and PDFs

These tools work with layers and masks in the Image Editor and with PDF documents on the board.

Tool	What it does
`inspect_image_layers`	Reads an image's layers and masks in the Image Editor, along with any active selected area and its preview or mask images.
`edit_image`	Adds, duplicates, reorders, renames, or removes Image Editor layers. Creates or removes layer and image masks, changes their visibility, locks, or opacity, and replaces layer or mask images.
`edit_pdf`	Creates a document from an imported PDF, navigates it, extracts or deletes pages, replaces its file, edits ink and text annotations, or exports the original or annotated PDF. For PDFs in table cells, it supports navigation and export.

To change how an e-book or comic is displayed, or move its reading position, the assistant uses edit_shape_text_and_content. PDF annotation editing requires a document on the board; a PDF in a table cell must first be moved out of the cell.

Animation

These tools read and build animations on the board. Reading or inspecting the animation pauses playback. The assistant can create up to 100 blocks or pauses at once.

Tool	What it does
`read_animation_timeline`	Reads a summary of the animation's sequences, blocks, and pauses, with filters by sequence, target shape, family, block kind, or time range.
`seek_animation_timeline`	Moves the animation playhead to a specific time and pauses playback so your assistant can inspect or capture that frame.
`inspect_animation_blocks`	Reads the full settings of specific animation blocks or pauses.
`create_animation_blocks`	Adds position, rotation, scale, resize, camera, opacity, stroke, text reveal, text sequence, video, audio, text-on-path crawl, attribute, or pause animations, with staggering where supported.
`edit_animation_blocks`	Changes the timing, target shapes, values, easing, or staggering of existing animation blocks and pauses.

Cameras, settings, and export

These tools change your view, cameras, and board or Player settings, and export snapshots of the board.

Tool	What it does
`move_viewport`	Moves or zooms your live view of the board to exact coordinates, to all visible content, or to specific shapes.
`edit_camera_settings`	Changes camera dimensions, resolution, zoom, lenses, and filter controls or code. Sets camera order, navigation, active-camera and passepartout options, or the Player's start and bounds cameras, playback range, sequence, and minimap.
`edit_board_settings`	Changes the board background, grid, grid image, reference axes, vignette, pivots, transformation gizmo, timeline duration, playback range and speed, animation base speeds, or snapping, tooltip, and language preferences.
`export_board`	Exports the current page, specific shapes, or a camera's view as PNG, JPEG, WebP, SVG, or shape JSON.

Undo and Versions

These tools work with the board's undo history and its Versions.

Tool	What it does
`undo_board_edits`	Undoes or redoes changes in the board's history, including your own edits.
`manage_board_versions`	Lists, creates, renames, deletes, or switches Versions. It can also reset chosen shapes or animation blocks to inherit from the main Version, or apply their changes to the main Version.

Use Versions for translations: keep the shared content in the main Version and translate the same shapes in a language Version.

Changes made with manage_board_versions cannot be undone, and switching Versions clears the undo history.

Helpers for your assistant

Your assistant can search and read the English documentation and built-in guides, then answer in your language. The documentation and icon tools work even without a connected board.

Tool	What it does
`search_docs`	Searches Irakasi's English documentation for pages and sections.
`read_docs`	Reads a page or section of Irakasi's English documentation and returns its link.
`read_guide`	Reads one of Irakasi's built-in guides and tool schemas for assistants.
`search_icons`	Searches Irakasi's icon library by name or keyword.
`read_icon_svg`	Reads the SVG drawing of an icon from the library so your assistant can inspect or place it.
`search_board_script_api`	Searches the board scripting helpers available to `run_board_script`.
`run_board_script`	Runs a short JavaScript function on the open board for tasks no dedicated tool covers. Its finished changes usually share one undo step; after a failure, your assistant checks the board.
`report_mcp_tool_problem`	Sends an anonymous report about a tool problem when problem reporting is enabled (see Privacy).