feishu-doc-reader – OpenClaw Skill

---

name: feishu-doc-reader description: Read and extract content from Feishu (Lark) documents using the official Feishu Open API metadata: {"moltbot":{"emoji":"📄","requires":{"bins":["python3","curl"]}}}

Feishu Document Reader

This skill enables reading and extracting content from Feishu (Lark) documents using the official Feishu Open API.

Configuration

Set Up the Skill

1. Create the configuration file at ./reference/feishu_config.json with your Feishu app credentials:

{
  "app_id": "your_feishu_app_id_here",
  "app_secret": "your_feishu_app_secret_here"
}

2. Make sure the scripts are executable:

chmod +x scripts/read_doc.sh
chmod +x scripts/get_blocks.sh

Security Note: The configuration file should be kept secure and not committed to version control. Consider using proper file permissions (chmod 600 ./reference/feishu_config.json).

Usage

Basic Document Reading

To read a Feishu document, you need the document token (found in the URL: https://example.feishu.cn/docx/DOC_TOKEN).

Using the shell script (recommended):

# Make sure environment variables are set first
./scripts/read_doc.sh "your_doc_token_here"

# Or specify document type explicitly
./scripts/read_doc.sh "docx_token" "doc"
./scripts/read_doc.sh "sheet_token" "sheet"

Get Detailed Document Blocks (NEW)

For complete document structure with all blocks, use the dedicated blocks script:

# Get full document blocks structure
./scripts/get_blocks.sh "docx_AbCdEfGhIjKlMnOpQrStUv"

# Get specific block by ID
./scripts/get_blocks.sh "docx_token" "block_id"

# Get blocks with children
./scripts/get_blocks.sh "docx_token" "" "true"

Using Python directly for blocks:

python scripts/get_feishu_doc_blocks.py --doc-token "your_doc_token_here"
python scripts/get_feishu_doc_blocks.py --doc-token "docx_token" --block-id "block_id"
python scripts/get_feishu_doc_blocks.py --doc-token "docx_token" --include-children

Supported Document Types

• Docx documents (new Feishu docs): Full content extraction with blocks, metadata, and structure
• Doc documents (legacy): Basic metadata and limited content
• Sheets: Full spreadsheet data extraction with sheet navigation
• Slides: Basic metadata (content extraction requires additional permissions)

Features

Enhanced Content Extraction

• Structured output: Clean JSON with document metadata, content blocks, and hierarchy
• Complete blocks access: Full access to all document blocks including text, tables, images, headings, lists, etc.
• Block hierarchy: Proper parent-child relationships between blocks
• Text extraction: Automatic text extraction from complex block structures
• Table support: Proper table parsing with row/column structure
• Image handling: Image URLs and metadata extraction
• Link resolution: Internal and external link extraction

Block Types Supported

• text: Plain text and rich text content
• heading1/2/3: Document headings with proper hierarchy
• bullet/ordered: List items with nesting support
• table: Complete table structures with cells and formatting
• image: Image blocks with tokens and metadata
• quote: Block quotes
• code: Code blocks with language detection
• equation: Mathematical equations
• divider: Horizontal dividers
• page: Page breaks (in multi-page documents)

Error Handling & Diagnostics

• Detailed error messages: Clear explanations for common issues
• Permission validation: Checks required permissions before making requests
• Token validation: Validates document tokens before processing
• Retry logic: Automatic retries for transient network errors
• Rate limiting: Handles API rate limits gracefully

Security Features

• Secure credential storage: Supports both environment variables and secure file storage
• No credential logging: Credentials never appear in logs or output
• Minimal permissions: Uses only required API permissions
• Access token caching: Efficient token reuse to minimize API calls

Command Line Options

Main Document Reader

# Python script options
python scripts/read_feishu_doc.py --help

# Shell script usage
./scripts/read_doc.sh <doc_token> [doc|sheet|slide]

Blocks Reader (NEW)

# Get full document blocks
./scripts/get_blocks.sh <doc_token>

# Get specific block
./scripts/get_blocks.sh <doc_token> <block_id>

# Include children blocks
./scripts/get_blocks.sh <doc_token> "" true

# Python options
python scripts/get_feishu_doc_blocks.py --help

API Permissions Required

Your Feishu app needs the following permissions:

• docx:document:readonly - Read document content
• doc:document:readonly - Read legacy document content
• sheets:spreadsheet:readonly - Read spreadsheet content

Error Handling

Common errors and solutions:

• 403 Forbidden: Check app permissions and document sharing settings
• 404 Not Found: Verify document token is correct and document exists
• Token expired: Access tokens are valid for 2 hours, refresh as needed
• App ID/Secret invalid: Double-check your credentials in Feishu Open Platform
• Insufficient permissions: Ensure your app has the required API permissions
• 99991663: Application doesn't have permission to access the document
• 99991664: Document doesn't exist or has been deleted
• 99991668: Token expired, need to refresh

Examples

Extract document with full structure

# Read document
./scripts/read_doc.sh "docx_AbCdEfGhIjKlMnOpQrStUv"

Get complete document blocks (NEW)

# Get all blocks with full structure
./scripts/get_blocks.sh "docx_AbCdEfGhIjKlMnOpQrStUv"

# Get specific block details
./scripts/get_blocks.sh "docx_AbCdEfGhIjKlMnOpQrStUv" "blk_xxxxxxxxxxxxxx"

Process spreadsheet data

./scripts/read_doc.sh "sheet_XyZ123AbCdEfGhIj" "sheet"

Extract only text content (Python script)

python scripts/read_feishu_doc.py --doc-token "docx_token" --extract-text-only

Security Notes

• Never commit credentials: Keep app secrets out of version control
• Use minimal permissions: Only request permissions your use case requires
• Secure file permissions: Set proper file permissions on secret files (chmod 600)
• Environment isolation: Use separate apps for development and production
• Audit access: Regularly review which documents your app can access

Troubleshooting

Authentication Issues

1. Verify your App ID and App Secret in Feishu Open Platform
2. Ensure the app has been published with required permissions
3. Check that environment variables or config files are properly set
4. Test with the test_auth.py script to verify credentials

Document Access Issues

1. Ensure the document is shared with your app or in an accessible space
2. Verify the document token format (should start with docx_, doc_, or sheet_)
3. Check if the document requires additional sharing permissions

Network Issues

1. Ensure your server can reach open.feishu.cn
2. Check firewall rules if running in restricted environments
3. The script includes retry logic for transient network failures

Blocks-Specific Issues

1. Empty blocks response: Document might be empty or have no accessible blocks
2. Missing block types: Some block types require additional permissions
3. Incomplete hierarchy: Use --include-children flag for complete block tree

References

• Feishu Open API Documentation
• Document API Reference
• Blocks API Reference
• Authentication Guide
• Sheet API Reference

Feishu Document Reader - Blocks Extraction

This skill provides comprehensive Feishu (Lark) document reading capabilities with full blocks extraction support.

Features

• Full blocks extraction: Get complete document structure including text, tables, images, headings, and more
• Multiple document types: Support for Docx (new), Doc (legacy), Sheets, and Slides
• Secure authentication: Proper token management with caching and refresh
• Error handling: Comprehensive error messages and diagnostics
• Easy integration: Simple command-line interface and Python API

Quick Start

1. Configuration

Create ./reference/feishu_config.json:

{
  "app_id": "your_feishu_app_id",
  "app_secret": "your_feishu_app_secret"
}

Set proper permissions:

chmod 600 ./reference/feishu_config.json

2. Usage

Get full document blocks (recommended):

# Using shell wrapper
./scripts/get_blocks.sh "docx_your_document_token"

# Using Python directly  
python scripts/get_feishu_doc_blocks.py --doc-token "docx_your_document_token" --output-format json

Get simplified text only:

python scripts/get_feishu_doc_blocks.py --doc-token "docx_your_document_token" --extract-text-only

3. Output Format

The full blocks output includes:

• document: Document metadata (title, revision, etc.)
• blocks: Complete block hierarchy with all content types
• text_content: Extracted plain text (when requested)

Integration with AI Agents

This skill can be used as a standalone tool or integrated into AI agent workflows:

1. Direct execution: Call the script from any AI agent
2. Extension tool: Register as an extension for seamless document processing
3. Pipeline integration: Combine with other tools for advanced document analysis

API Permissions Required

Your Feishu app needs these permissions in Open Platform:

• docx:document:readonly (for new documents)
• doc:document:readonly (for legacy documents)
• sheets:spreadsheet:readonly (for spreadsheets)

Security Notes

• Credentials are never logged or exposed in output
• Access tokens are cached and refreshed automatically
• File system access is restricted to prevent path traversal
• Use minimal required permissions for your use case

Troubleshooting

Common Issues

Authentication Failed (401):

• Verify App ID and App Secret in Feishu Open Platform
• Ensure app is published with required permissions

Document Not Found (404):

• Check document token format (should start with docx_, doc_, or sheet_)
• Ensure document is shared with your app

Permission Denied (403):

• Verify required API permissions are granted
• Check if document requires additional sharing settings

Debugging

Enable debug logging:

DEBUG=1 python scripts/get_feishu_doc_blocks.py --doc-token "your_token"

Examples

See USAGE_EXAMPLES.md for detailed examples.

References

• Feishu Open API Documentation
• Document Blocks API
• Authentication Guide