What problem does it solve? Working with Feishu (Lark) online spreadsheets through raw APIs is error-prone: agents guess command names, misplace sheets, write past the real data range, or fake charts and pivot tables with static values. This Skill provides a verified command map and strict workflow rules so spreadsheet operations are executed correctly against the live document. ## Core Features & Use Cases - Full spreadsheet CRUD: create workbooks, manage sheets, insert/delete/resize/freeze rows and columns, read and write cell values, formulas, styles, comments, and cell images. - Data objects: create and maintain charts, pivot tables, conditional formats, filters, filter views, sparklines, and floating images as real spreadsheet objects that update with source data. - Atomic batch updates: combine multiple write operations into a single transactional +batch-update call with dry-run preview and mandatory read-back verification. - Use Case: A user asks to group sales data by region and chart the result. The Skill reads the workbook structure, creates a real pivot table on a new sheet, excludes the grand-total row, and builds a column chart referencing the exact data range. ## Quick Start Ask the AI to read the Feishu spreadsheet at your /sheets/ URL and summarize the data in its first worksheet.
