lark-whiteboard

Query, export, and edit Feishu whiteboards using lark-cli and whiteboard-cli.

8|2|Updated Apr 29, 2026
One-click install
npx skills add https://github.com/luojiaaoo/xpeech --skill lark-whiteboard-luojiaaoo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: lark-whiteboard
Source: https://github.com/luojiaaoo/xpeech/tree/main/xpeech/agent/skills/buildin/lark-skills/lark-whiteboard
Command: npx skills add https://github.com/luojiaaoo/xpeech --skill lark-whiteboard-luojiaaoo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @larksuite/whiteboard-cli, and includes references (resource) components.

What problem does it solve? Creating and editing diagrams in Feishu (Lark) whiteboards manually is slow and hard to automate. This Skill lets an agent read whiteboard content, export it as preview images, SVG, or Mermaid/PlantUML source, and write new or updated diagrams back to the board programmatically. ## Core Features & Use Cases - Export whiteboards: Export a board as a preview image, SVG vector graphic, Mermaid/PlantUML source code, or raw OpenAPI node structure via lark-cli whiteboard +export. - Create and update diagrams: Write content with +update in Mermaid, PlantUML, SVG, or raw OpenAPI formats, with overwrite or incremental append modes and idempotent tokens for safe retries. - Rich diagram generation: Built-in scene guides and DSL layout rules for architecture diagrams, flowcharts, fishbone diagrams, funnels, flywheels, bar/line charts, comparison matrices, and more, rendered locally with @larksuite/whiteboard-cli before writing. - Use Case: Ask the agent to draw a microservices architecture diagram into a Feishu doc's whiteboard; it generates the DSL JSON, renders a PNG preview for verification, converts to OpenAPI format, and writes it to the board. ## Quick Start Use the lark-whiteboard skill to export the whiteboard at this Feishu doc URL as a preview image, then create a new architecture diagram and write it back to the board.

Frequently Asked Questions about lark-whiteboard

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I export a Feishu whiteboard as an image or SVG?▼

Run lark-cli whiteboard +export with the board token and --output-type set to preview, svg, source, or raw. Preview saves an image file, svg exports a vector graphic, and source extracts Mermaid or PlantUML code when the board contains exactly one such diagram.

How do I create a diagram in a Feishu whiteboard from Mermaid or PlantUML code?▼

Write your Mermaid or PlantUML code to a file, then run lark-cli whiteboard +update with --input_format mermaid or plantuml and --source pointing to the file or stdin. Add --overwrite to replace existing board content or omit it to append.

What diagram types can be generated for Feishu whiteboards?▼

The skill supports architecture diagrams, flowcharts, organization charts, swimlanes, comparison tables, fishbone diagrams, bar and line charts, treemaps, funnels, pyramids, flywheels, and milestones. Complex geometric diagrams are generated via CommonJS scripts that compute coordinates and emit DSL JSON.

Can I edit an existing Feishu whiteboard without losing its content?▼

Yes, use incremental append mode by omitting --overwrite so new nodes are added while original content stays untouched. For structural changes, the SVG edit route or source regeneration is used, but overwrite modes rebuild the whole board and require user confirmation first.

Why does adding images to a Feishu whiteboard fail with a 500 error?▼

Whiteboard image nodes require a media token from the whiteboard domain. Download the image, then upload it with lark-cli docs +media-upload --parent-type whiteboard --parent-node <board_token>, and reference the returned media token in the DSL image node.

What tools are required to use Feishu whiteboard automation?▼

You need lark-cli for authenticated Feishu API operations and the npm package @larksuite/whiteboard-cli for rendering DSL JSON to PNG and converting it to OpenAPI format. Both are verified at skill startup before any board operation.