Feishu Document Tool

Single tool feishu_doc with action parameter for all document operations, including table creation for Docx.

Token Extraction

From URL https://xxx.feishu.cn/docx/ABC123def → doc_token = ABC123def

Actions

Read Document

{ "action": "read", "doc_token": "ABC123def" }

Returns: title, plain text content, block statistics. Check hint field — if present, structured content (tables, images) exists that requires list_blocks.

Write Document (Replace All)

{ "action": "write", "doc_token": "ABC123def", "content": "# Title\
\
Markdown content..." }

Replaces entire document with markdown content. Supports: headings, lists, code blocks, quotes, links, images (![](url) auto-uploaded), bold/italic/strikethrough, and Markdown tables.

Append Content

{ "action": "append", "doc_token": "ABC123def", "content": "Additional content" }

Appends markdown to end of document. Supports the same content types as write, including Markdown tables.

Insert Content (Positioned)

{
  "action": "insert",
  "doc_token": "ABC123def",
  "after_block_id": "doxcnXXX",
  "content": "Markdown content including tables..."
}

Inserts content after a specific block. Uses Descendant API internally, supports all block types including tables.

Key usage: This is the primary way to insert content at a precise position. Use list_blocks first to find the after_block_id.

Create Document

{ "action": "create", "title": "New Document", "owner_open_id": "ou_xxx" }

With folder:

{
  "action": "create",
  "title": "New Document",
  "folder_token": "fldcnXXX",
  "owner_open_id": "ou_xxx"
}

Important: Always pass owner_open_id with the requesting user's open_id (from inbound metadata sender_id) so the user automatically gets full_access permission on the created document. Without this, only the bot app has access.

List Blocks

{ "action": "list_blocks", "doc_token": "ABC123def" }

Returns full block data including tables, images. Use this to read structured content and find block IDs for positioning.

Get Single Block

{ "action": "get_block", "doc_token": "ABC123def", "block_id": "doxcnXXX" }

Update Block Text

{
  "action": "update_block",
  "doc_token": "ABC123def",
  "block_id": "doxcnXXX",
  "content": "New text with **bold** and `code`"
}

Supports Markdown inline styles: **bold**, `code`, [link](url), ~~strike~~.

Supported block types: Text, Heading, Bullet, Ordered, Code, Quote, Todo.

⚠️ NOT for tables: Using update_block on a table cell replaces content as plain text, destroying all inline formatting (inline_code, bold, etc.).

Delete Block

{ "action": "delete_block", "doc_token": "ABC123def", "block_id": "doxcnXXX" }

Create Table

{
  "action": "create_table",
  "doc_token": "ABC123def",
  "row_size": 2,
  "column_size": 2,
  "column_width": [200, 200]
}

⚠️ Positioning unreliable: parent_block_id and index parameters do not work correctly for tables — tables always end up at the document end. Use insert action with after_block_id instead for positioned tables.

Write Table Cells

{
  "action": "write_table_cells",
  "doc_token": "ABC123def",
  "table_block_id": "doxcnTABLE",
  "values": [
    ["A1", "B1"],
    ["A2", "B2"]
  ]
}

Writes plain text into table cells. Supports Markdown inline styles in cell values.

Note: This clears and rebuilds cell children. Suitable for initial table filling or full-cell replacement, but not for partial cell edits.

Create Table With Values (One-step)

{
  "action": "create_table_with_values",
  "doc_token": "ABC123def",
  "row_size": 2,
  "column_size": 2,
  "column_width": [200, 200],
  "values": [["A1", "B1"], ["A2", "B2"]]
}

⚠️ Same positioning caveat as create_table — use insert action for positioned tables.

Table Row/Column Operations

{ "action": "insert_table_row", "doc_token": "...", "table_block_id": "...", "row_index": -1 }
{ "action": "insert_table_column", "doc_token": "...", "table_block_id": "...", "column_index": -1 }
{ "action": "delete_table_rows", "doc_token": "...", "block_id": "...", "row_start": 1, "row_count": 1 }
{ "action": "delete_table_columns", "doc_token": "...", "block_id": "...", "column_start": 0, "column_count": 1 }
{ "action": "merge_table_cells", "doc_token": "...", "table_block_id": "...", "row_start": 0, "row_end": 2, "column_start": 0, "column_end": 2 }

Note for delete_table_rows: Uses block_id (not table_block_id), row_start (not row_index).

Upload Image

{ "action": "upload_image", "doc_token": "ABC123def", "url": "https://example.com/image.png" }

Or local path with position control:

{
  "action": "upload_image",
  "doc_token": "ABC123def",
  "file_path": "/tmp/image.png",
  "parent_block_id": "doxcnParent",
  "index": 5
}

Or base64 data URI:

{
  "action": "upload_image",
  "doc_token": "ABC123def",
  "image": "data:image/png;base64,iVBOR..."
}

Note: Image display size is determined by the uploaded image's pixel dimensions. For small images (e.g. 480x270), scale to 800px+ width before uploading.

Upload File Attachment

{ "action": "upload_file", "doc_token": "ABC123def", "url": "https://example.com/report.pdf" }

Or local path:

{ "action": "upload_file", "doc_token": "ABC123def", "file_path": "/tmp/report.pdf", "filename": "Q1-report.pdf" }

Color Text

{
  "action": "color_text",
  "doc_token": "ABC123def",
  "block_id": "doxcnXXX",
  "content": "colored text",
  "color": "red"
}

Reading Workflow

1. Start with action: "read" — get plain text + statistics
2. Check block_types in response for Table, Image, Code, etc.
3. If structured content exists, use action: "list_blocks" for full data

Incremental Sync Best Practices

When to Use Incremental vs Full Sync

• < 5 blocks changed → incremental (delete + insert individual blocks)
• > 80% blocks changed → full sync (write action to replace all)
• Single table changed → delete old table + insert new table

Table Update Strategy: Delete + Insert

Tables should NEVER be edited in-place. Always:

1. list_blocks → find the block before the table (anchor_block_id)
2. delete_block(block_id = old_table_id)
3. insert(after_block_id = anchor_block_id, content = "| new | table |...")

Why not edit in-place?
- update_block on table cell → strips inline formatting (inline_code, bold, links all become plain text)
- write_table_cells → clears and rebuilds cell children, similar formatting loss
- create_table / create_table_with_values → positioning unreliable

Multi-Block Batch Replacement

When changes span multiple consecutive blocks:

1. Find anchor (last unchanged block before the changed region)
2. Delete changed blocks in reverse order (to avoid index drift)
3. Single insert call with all new content (supports mixed block types)

Diff Strategy (from feishu-doc-sync)

1. Block signature matching — (type, content) hash for alignment, style hash for change detection
2. LCS alignment of local ↔ remote blocks
3. Three-phase execution: UPDATE (text blocks) → DELETE (reverse order) → INSERT (forward order)
4. Tables: if structure is identical, cell-level diff is possible; if structure differs, full table replacement
5. When change rate > 80%, fall back to full sync

Table Header Row (Dark Background)

The header_row: true property gives tables a dark header background. This is a common requirement.

⚠️ Critical: header_row can ONLY be set at creation time

The PATCH API (update_table_property) does NOT work for header_row — it returns code: 0 (success) but silently fails to apply the property. This applies to both single-block PATCH and batch_update.

How to Create Tables with Header Row

Since insert action (Descendant API) and create_table action do not support header_row, you must use the raw Feishu API:

# Step 1: Get tenant_access_token
TOKEN=$(curl -s https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal \
  -H "Content-Type: application/json" \
  -d '{"app_id":"'$APP_ID'","app_secret":"'$APP_SECRET'"}' \
  | python3 -c "import json,sys; print(json.load(sys.stdin)['tenant_access_token'])")

# Step 2: Create empty table with header_row via document_block_children.create
curl -s -X POST \
  "https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks/{doc_token}/children?document_revision_id=-1" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "children": [{
      "block_type": 31,
      "table": {
        "property": {
          "row_size": 3,
          "column_size": 2,
          "header_row": true,
          "column_width": [300, 400]
        }
      }
    }],
    "index": 5
  }'

# Step 3: Fill cells via feishu_doc write_table_cells
# (use the new table's block_id from step 2 response)

Credentials: Read from /root/.openclaw/openclaw.json → channels.feishu.appId / channels.feishu.appSecret.

Workflow for Positioned Table with Header Row

1. list_blocks → find the anchor block and target index
2. (If replacing) delete_block → remove old table
3. Raw API document_block_children.create → create empty table with header_row: true at correct index
4. feishu_doc write_table_cells → fill content
5. Verify: get_block → check table.property.header_row

Column Width Algorithm

Auto-calculated when using insert action (Descendant API):
- Total width: ~730px
- CJK characters count as 2x width
- Per-column: proportional distribution, clamped to 50–400px
- When using raw API, specify column_width manually

Sync Prerequisites

• Always git pull before reading local repo files — ensures latest content
• Compare before writing — use read or list_blocks to check current state
• Preserve formatting — inline_code, bold, links in table cells require careful handling

Configuration

channels:
  feishu:
    tools:
      doc: true  # default: true

Note: feishu_wiki depends on this tool — wiki page content is read/written via feishu_doc.

Permissions

Required: docx:document, docx:document:readonly, docx:document.block:convert, drive:drive