The Whimsical MCP gives AI assistants like Claude, Cursor, and Codex a set of tools to read and write content in your Whimsical workspace. These tools work together. For example, a typical session might search for an existing file, read its content, then create or edit a new diagram based on what it finds.

This article lists the MCP tools currently available.

MCP toolsCopied!Link to this section

searchCopied!Link to this section

Search across your Whimsical workspace by title or content.

Use this to find existing boards, docs, folders, and sections before creating something new, or to pull relevant files into context. Supports filtering to file and folder titles only, or searching across all content, including shapes and text on boards.

Example prompts:

• "Find the system architecture diagram from last quarter"
• "Search for files about the onboarding flow"
• "Do we have any wireframes for the settings page?"

fetchCopied!Link to this section

Read the contents of a Whimsical board, doc, or folder by ID or URL.

Returns the text content of the file. For boards, this includes all shapes, sticky notes, text, and diagram nodes. For docs, it returns the full text. Can also return a PNG image snapshot of a board. Use search first to find the file ID.

Example prompts:

• "Read the user flow diagram and summarise the key steps"
• "What does our current API architecture diagram show?"
• "Take a screenshot of that wireframe and describe what's in it"

createCopied!Link to this section

Create a new Whimsical file — a board, flowchart, mind map, sequence diagram, wireframe, or sticky note set.

This is the primary tool for generating new visual content. Diagrams are auto-laid-out; boards support freeform placement. New files are created in your workspace, and a link is returned.

Example prompts:

• "Create a flowchart of the OAuth authentication flow"
• "Generate a mind map of the key features we're building this quarter"
• "Make a sequence diagram showing how an API request flows through our system"
• "Build a wireframe for a new settings page with a sidebar and form fields"
• "Create a board visualising the relationships between our microservices"
• "Create a table comparing the three pricing tiers"
• "Add some sticky notes brainstorming ideas for the new feature"

Supported file types:

• flowchart — Process flows and decision trees, auto-laid-out
• mindmap — Hierarchical idea maps, auto-laid-out
• sequence_diagram — Interaction diagrams between actors, auto-laid-out
• wireframe — UI mockups with flexbox layout
• board — Freeform canvas with full positional control
• sticky_notes — Clusters of sticky notes, auto-laid-out
• table — Structured data tables from markdown or column/row data
• stamp — Visual stamps and badges

doc_createCopied!Link to this section

Create a new Whimsical document with markdown content.

Supports headings, lists, links, code blocks, and tables. Use this for text-heavy content rather than create.

Example prompts:

• "Create a doc outlining the technical spec for the new API"
• "Write up the meeting notes as a Whimsical doc"
• "Make a document listing the pros and cons of each approach"

editCopied!Link to this section

Add, update, or delete objects on an existing Whimsical board or doc.

For boards: add shapes, text, sticky notes, connectors, icons, and tables; update position, color, size, or text; delete objects. For docs: insert, update, or delete blocks. Also supports find-and-replace across a diagram, and structural updates for mind maps and sequence diagrams.

Example prompts:

• "Add a 'Rate Limiting' step to the API flow diagram"
• "Update the status on that flowchart from 'Draft' to 'Approved'"
• "Add a new section to the architecture doc about caching"
• "Remove the deprecated service from the microservices diagram"
• "Replace all instances of 'User' with 'Customer' in the diagram"

auto_layoutCopied!Link to this section

Re-arrange a flowchart using automatic layout.

Useful after adding or editing shapes to clean up positioning. Connectors re-route automatically. Supports top-to-bottom, left-to-right, and other orientations, as well as compact and default spacing.

Example prompts:

• "Tidy up the layout of that flowchart after the changes"
• "Re-arrange the flowchart left-to-right instead of top-to-bottom"

file_treeCopied!Link to this section

Browse the folder hierarchy in your workspace.

Use this to see how files are organized, find a specific folder to create new content in, or get an overview of a project area.

Example prompts:

• "Show me the folder structure of my workspace"
• "What's inside the Q3 Planning folder?"
• "Where are the design files for the mobile app?"

list_workspacesCopied!Link to this section

List all Whimsical workspaces you belong to.

Returns workspace names, IDs, and your role (member or guest). Useful when you have access to multiple workspaces and need to target the right one.

Example prompts:

• "Which Whimsical workspaces do I have access to?"
• "Switch to the Design Team workspace"

comment_readCopied!Link to this section

Read the comment threads on a Whimsical board or doc.

Returns all threads with their author, timestamp, and content. Can read all comments on a file, or filter to comments on a specific object or block. Useful for pulling design feedback or decisions into context before making edits.

Example prompts:

• "What feedback is there on the checkout wireframe?"
• "Summarise the comments on the architecture board"
• "Are there any unresolved comments on this doc?"

comment_editCopied!Link to this section

Create, reply to, edit, resolve, reopen, or delete a comment thread on a Whimsical board or doc.

Useful for leaving notes, recording decisions, or flagging questions directly on a file. Comments support markdown formatting.

Example prompts:

• "Leave a comment on the wireframe asking about the empty state"
• "Reply to the thread about the login flow with a summary of what we decided"
• "Mark that comment thread as resolved"
• "Reopen the comment about the navigation layout"

How the tools work togetherCopied!Link to this section

Most tasks combine several tools in sequence. For example:

"Update our architecture diagram to add the new caching layer"

1. search — finds the architecture board
2. fetch — reads the current diagram content
3. edit — adds the new caching layer node and connectors
4. auto_layout — tidies up the positions

"Summarise the feedback on our latest wireframes"

1. search — finds the wireframe files
2. comment_read — retrieves all comment threads
3. AI summarises the feedback in the chat

"Turn this codebase into a system architecture diagram"

1. AI reads the codebase (via its own tools)
2. create — generates a new board or flowchart in Whimsical
3. Link returned in chat for review

"Create a spec doc in the right project folder"

1. file_tree — browses the workspace to find the target folder
2. doc_create — creates the document in that folder

Tips for better promptsCopied!Link to this section

• Be specific about the diagram type. "Create a flowchart" or "create a sequence diagram" gets better results than "make a diagram".
• Reference existing files by name. "Update the authentication flowchart" gives the AI enough to search for the right file.
• Describe the content, not just the task. "Create a mind map of our Q3 initiatives: Discovery, Payments, and Growth, with sub-topics for each" produces richer output than "create a mind map".
• Ask it what it can do. Prompt "What can you do with Whimsical MCP?" at the start of a session to see a summary of available tools.