What problem does it solve? Managing Feishu (Lark) cloud documents from the terminal or an AI agent normally requires juggling raw OpenAPI calls, token types, and wiki link resolution. This Skill wraps those operations into high-level lark-cli shortcuts so you can create, read, update, and search documents without hand-writing API requests. ## Core Features & Use Cases - Document lifecycle management: Create documents from Lark-flavored Markdown, fetch content as Markdown, and update with 7 modes (append, replace_range, replace_all, insert_before, insert_after, delete_range, overwrite). - Media handling: Insert local images/files into documents, preview or download embedded media, and download whiteboard thumbnails. - Cloud-space search: Use Search v2 to discover docs, Wiki nodes, spreadsheets, and bitables with Boolean query syntax and rich filters (owner, type, time range, folder, chat). - Use Case: Ask the agent to "create a project plan doc in my wiki space with an architecture whiteboard" — it creates the doc, inserts a blank whiteboard, and hands off to the whiteboard skill to fill in the diagram. ## Quick Start Ask the agent to create a Feishu document titled "Weekly Report" from a Markdown outline, or to fetch the content of a given Lark document URL.