feishu-doc

The Feishu Doc agent skill provides deep integration for the Feishu/Lark document ecosystem, allowing your AI assistant to act as a collaborative partner for documentation management. It enables automated interaction with Feishu Docx, facilitating tasks such as content creation, document summarization, and structural manipulation of complex blocks. Whether you are drafting meeting minutes, managing project requirements in cloud docs, or programmatically injecting data into tables, this tool serves as the primary interface for Feishu document operations.

• Full support for reading plain text, counting block statistics, and listing complex document structures like images and code blocks.

• Capabilities to create new documents, append markdown content, or replace entire documents with generated text.

• Advanced block-level operations including updating text, deleting specific segments, and creating dynamic tables with custom widths.

• Media handling features that allow uploading images and file attachments directly to a doc via URL or local path.

• Automated permission management by accepting owner_open_id to ensure the bot-created content remains accessible to the user.

• Always extract the doc_token from Feishu URLs (e.g., https://xxx.feishu.cn/docx/ABC123def) to initiate interactions.

• For structured documents containing tables or images, use the read action first to check hints, then invoke list_blocks for full data resolution.

• When creating new documents, ensure the owner_open_id parameter is passed from the sender_id metadata to prevent permission locked-out scenarios.

• Note that standard markdown tables are not supported; use the create_table_with_values action for structured table generation.

• Image dimensions should ideally exceed 800px width for optimal display, especially for smaller assets like GIFs.

• The tool integrates natively with feishu_wiki, meaning wiki page content is also managed through these document-specific functions.