hellostar999

---
name: feishu-ops
description: |
  飞书（Feishu/Lark）文档与消息操作技能。When to use: 用户要求创建、删除、修改飞书文档；查询或更新文档中指定行/列的数据；向飞书联系人或群聊发送消息。Triggers: "创建飞书文档"、"删除文档"、"修改文档内容"、"更新第X行第Y列"、"查询文档"、"发送飞书消息"、"发消息给群"。
---

# 飞书操作技能 (feishu-ops)

支持飞书文档的完整 CRUD 以及单元格精细操作，同时支持发送消息给个人和群聊。

## 凭证配置

所有操作需要 `app_id` 和 `app_secret`，位于技能目录的 `scripts/config.json`：

```json
{
  "app_id": "cli_xxxxx",
  "app_secret": "xxxxx"
}
```

若无配置文件，脚本会报错并提示创建。

## 文档操作

### 创建文档

```python
python scripts/feishu_doc.py create "文档标题"
```

成功返回：`{"doc_token": "xxx", "doc_url": "https://feishu.cn/doc/xxx"}`

### 删除文档

```python
python scripts/feishu_doc.py delete <doc_token>
```

### 写入/追加文档内容

```python
# 全量写入（覆盖）
python scripts/feishu_doc.py write <doc_token> "## 标题\n这是内容"

# 追加段落
python scripts/feishu_doc.py append <doc_token> "新追加的段落"
```

### 读取文档

```python
python scripts/feishu_doc.py read <doc_token>
```

### 精细操作：单元格读写

飞书文档的 blocks API 支持按 block_id 精确操作。使用前需先 `read` 文档获取 block 树结构。

```python
# 查询指定 block 内容
python scripts/feishu_doc.py query-block <doc_token> <block_id>

# 更新指定 block 的文本内容
python scripts/feishu_doc.py update-block <doc_token> <block_id> "新的文本内容"

# 在文档末尾追加一个段落 block
python scripts/feishu_doc.py append-block <doc_token> "段落文本"
```

> **表格操作**：表格为复合 block。先用 `read` 获取表格 block_id，再遍历表格内部 cell blocks 进行读写。

## 消息操作

### 发送文本消息给用户

需要目标用户的 `open_id`（可用 `search_user.py` 查询）。

```python
python scripts/feishu_msg.py send-user <open_id> "消息内容"
```

### 发送文本消息到群

需要目标群的 `chat_id`（可用 `search_chat.py` 查询）。

```python
python scripts/feishu_msg.py send-chat <chat_id> "消息内容"
```

### 查询群聊消息

```python
python scripts/feishu_msg.py get-messages <chat_id> [page_size]
```

返回群内消息列表，自动正确显示中文内容。

### 发送文件到群

需要先安装 SDK：
```bash
pip install lark-oapi
```

发送本地文件到群：
```python
python scripts/feishu_msg.py send-file <chat_id> <本地文件路径>
```

例如发送到龙虾测试群：
```python
python scripts/feishu_msg.py send-file oc_2c6df8f6e06e88d34729baacc124b89e "C:\\Users\\10430\\Desktop\\采购数据.xlsx"
```

### 查询用户 open_id

```python
python scripts/feishu_msg.py search-user <姓名关键词>
```

### 查询群聊 chat_id

```python
python scripts/feishu_msg.py search-chat <群名关键词>
```

## 常用工作流

**创建文档并写入内容：**
1. `create` 获取 doc_token
2. `write` 或 `append-block` 写入内容
3. `read` 确认内容

**更新表格中第R行第C列：**
1. `read` 获取文档 block 树
2. 找到目标表格的 block_id
3. 遍历表格 rows/cells，用 `update-block` 更新目标单元格

**发消息给同事：**
1. `search-user` 查找 open_id
2. `send-user` 发送消息

## 脚本索引

| 脚本 | 功能 |
|------|------|
| `scripts/feishu_doc.py` | 文档 CRUD + block 精细操作 |
| `scripts/feishu_msg.py` | 消息发送 + 用户/群查询 + 获取消息 + 文件发送（需 lark-oapi SDK） |
| `scripts/config.json` | 凭证配置 |
| `references/api_ref.md` | 完整 API 参数说明 |

FILE:scripts/config.json
{
  "app_id": "",
  "app_secret": ""
}

FILE:scripts/feishu_doc.py
"""
飞书文档操作脚本
用法:
  python feishu_doc.py create "<标题>"
  python feishu_doc.py delete <doc_token>
  python feishu_doc.py read <doc_token>
  python feishu_doc.py write <doc_token> "<markdown内容>"
  python feishu_doc.py append <doc_token> "<文本>"
  python feishu_doc.py append-block <doc_token> "<段落文本>"
  python feishu_doc.py query-block <doc_token> <block_id>
  python feishu_doc.py update-block <doc_token> <block_id> "<新文本>"
"""

import json
import sys
import urllib.request
import urllib.error

SCRIPT_DIR = __file__.rsplit("/", 1)[0] if "/" in __file__ else __file__.rsplit("\\", 1)[0]
CONFIG_PATH = SCRIPT_DIR + "/config.json"


def load_config():
    with open(CONFIG_PATH, "r", encoding="utf-8") as f:
        return json.load(f)


def get_tenant_access_token(app_id, app_secret):
    url = "https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal"
    data = json.dumps({"app_id": app_id, "app_secret": app_secret}).encode("utf-8")
    req = urllib.request.Request(url, data=data, headers={"Content-Type": "application/json"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    return result.get("tenant_access_token", "")


def create_document(title):
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    url = "https://open.feishu.cn/open-apis/docx/v1/documents"
    data = json.dumps({"title": title}).encode("utf-8")
    req = urllib.request.Request(url, data=data, headers={
        "Content-Type": "application/json",
        "Authorization": f"Bearer {token}"
    })
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    if result.get("code") != 0:
        print(json.dumps(result, ensure_ascii=False))
        return
    doc = result["data"]["document"]
    print(json.dumps({"doc_token": doc["document_id"], "doc_url": f"https://feishu.cn/docx/{doc['document_id']}"}, ensure_ascii=False))


def delete_document(doc_token):
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    url = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}"
    req = urllib.request.Request(url, method="DELETE", headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    print(json.dumps(result, ensure_ascii=False))


def read_document(doc_token):
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    url = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks?page_size=500"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    print(json.dumps(result, ensure_ascii=False, indent=2))


def _text_to_blocks(text):
    """将文本转为 Text Block（block_type=2）"""
    blocks = []
    for line in text.split("\n"):
        blocks.append({
            "block_type": 2,  # Text Block
            "text": {
                "elements": [{"type": "text_run", "text_run": {"content": line}}]
            }
        })
    return blocks


def write_document(doc_token, content):
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])

    # 获取现有 blocks
    url = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks?page_size=500"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())

    if result.get("code") != 0:
        print(json.dumps(result, ensure_ascii=False))
        return

    items = result["data"]["items"]
    if not items:
        print(json.dumps({"error": "文档为空或不存在"}))
        return

    root_block_id = items[0]["block_id"]

    # 写入：新 blocks
    url2 = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks/{root_block_id}/children"
    blocks = _text_to_blocks(content)
    data = json.dumps({"children": blocks, "index": 0}).encode("utf-8")
    req2 = urllib.request.Request(url2, data=data, method="POST", headers={
        "Content-Type": "application/json; charset=utf-8",
        "Authorization": f"Bearer {token}"
    })
    with urllib.request.urlopen(req2) as resp:
        result2 = json.loads(resp.read())
    print(json.dumps(result2, ensure_ascii=False))


def append_paragraph(doc_token, text):
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])

    url = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks?page_size=500"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())

    if result.get("code") != 0:
        print(json.dumps(result, ensure_ascii=False))
        return

    items = result["data"]["items"]
    if not items:
        print(json.dumps({"error": "文档为空"}))
        return

    root_block_id = items[0]["block_id"]

    url2 = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks/{root_block_id}/children"
    blocks = _text_to_blocks(text)
    data = json.dumps({"children": blocks}).encode("utf-8")
    req2 = urllib.request.Request(url2, data=data, method="POST", headers={
        "Content-Type": "application/json; charset=utf-8",
        "Authorization": f"Bearer {token}"
    })
    with urllib.request.urlopen(req2) as resp:
        result2 = json.loads(resp.read())
    print(json.dumps(result2, ensure_ascii=False))


def query_block(doc_token, block_id):
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    url = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks/{block_id}"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    print(json.dumps(result, ensure_ascii=False, indent=2))


def update_block_text(doc_token, block_id, new_text):
    """更新指定 block 的文本内容（仅支持 Text/Heading/Bullet 等文本类 block）"""
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    url = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_token}/blocks/{block_id}"
    data = json.dumps({
        "update_text_elements": {
            "elements": [{"type": "text_run", "text_run": {"content": new_text}}]
        }
    }).encode("utf-8")
    req = urllib.request.Request(url, data=data, method="PATCH", headers={
        "Content-Type": "application/json; charset=utf-8",
        "Authorization": f"Bearer {token}"
    })
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    print(json.dumps(result, ensure_ascii=False))


def append_block(doc_token, text):
    """追加段落 block（便捷别名）"""
    append_paragraph(doc_token, text)


if __name__ == "__main__":
    args = sys.argv[1:]
    if len(args) < 1:
        print("用法: python feishu_doc.py <command> [args...]")
        sys.exit(1)

    cmd = args[0]

    if cmd == "create" and len(args) >= 2:
        create_document(args[1])
    elif cmd == "delete" and len(args) >= 2:
        delete_document(args[1])
    elif cmd == "read" and len(args) >= 2:
        read_document(args[1])
    elif cmd == "write" and len(args) >= 3:
        write_document(args[1], args[2])
    elif cmd == "append" and len(args) >= 3:
        append_paragraph(args[1], args[2])
    elif cmd == "append-block" and len(args) >= 3:
        append_block(args[1], args[2])
    elif cmd == "query-block" and len(args) >= 3:
        query_block(args[1], args[2])
    elif cmd == "update-block" and len(args) >= 4:
        update_block_text(args[1], args[2], args[3])
    else:
        print("未知命令或参数不足")
        sys.exit(1)

FILE:scripts/feishu_msg.py
"""
飞书消息操作脚本（支持文件发送）
用法:
  python feishu_msg.py send-user <open_id> <消息内容>
  python feishu_msg.py send-chat <chat_id> <消息内容>
  python feishu_msg.py search-user <姓名关键词>
  python feishu_msg.py search-chat <群名关键词>
  python feishu_msg.py get-messages <chat_id> [page_size]
  python feishu_msg.py send-file <chat_id> <本地文件路径>
"""

import json
import sys
import urllib.parse
import uuid
import datetime
import os

# 全局 UTF-8 输出修复（Windows GBK 终端适配）
sys.stdout.reconfigure(encoding='utf-8', errors='replace')

SCRIPT_DIR = __file__.rsplit("/", 1)[0] if "/" in __file__ else __file__.rsplit("\\", 1)[0]
CONFIG_PATH = SCRIPT_DIR + "/config.json"

# 尝试导入 lark-oapi SDK，upload_file 函数会用到
try:
    import lark_oapi as lark
    from lark_oapi.api.im.v1 import *
    SDK_AVAILABLE = True
except ImportError:
    SDK_AVAILABLE = False


def load_config():
    with open(CONFIG_PATH, "r", encoding="utf-8") as f:
        return json.load(f)


def get_tenant_access_token(app_id, app_secret):
    url = "https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal"
    data = json.dumps({"app_id": app_id, "app_secret": app_secret}).encode("utf-8")
    req = urllib.request.Request(url, data=data, headers={"Content-Type": "application/json"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    return result.get("tenant_access_token", "")


def send_message(receive_id, receive_id_type, content, msg_type="text"):
    if not SDK_AVAILABLE:
        # fallback to urllib
        import urllib.request
        config = load_config()
        token = get_tenant_access_token(config["app_id"], config["app_secret"])
        url = "https://open.feishu.cn/open-apis/im/v1/messages?receive_id_type=" + receive_id_type
        payload = {
            "receive_id": receive_id,
            "msg_type": msg_type,
            "content": json.dumps({"text": content}) if msg_type == "text" else content,
            "uuid": str(uuid.uuid4())
        }
        data = json.dumps(payload).encode("utf-8")
        req = urllib.request.Request(url, data=data, method="POST", headers={
            "Content-Type": "application/json; charset=utf-8",
            "Authorization": f"Bearer {token}"
        })
        with urllib.request.urlopen(req) as resp:
            result = json.loads(resp.read())
        print(json.dumps(result, ensure_ascii=False))
        return result

    # Use SDK
    client = lark.Client.builder() \
        .app_id(load_config()["app_id"]) \
        .app_secret(load_config()["app_secret"]) \
        .log_level(lark.LogLevel.ERROR) \
        .build()

    request = CreateMessageRequest.builder() \
        .receive_id_type(receive_id_type) \
        .request_body(CreateMessageRequestBody.builder()
            .receive_id(receive_id)
            .msg_type(msg_type)
            .content(content if msg_type != "text" else json.dumps({"text": content}))
            .build()) \
        .build()

    response = client.im.v1.message.create(request)
    print(json.dumps({"code": response.code, "msg": response.msg}, ensure_ascii=False))
    return {"code": response.code, "msg": response.msg}


def send_user(open_id, content):
    return send_message(open_id, "open_id", content)


def send_chat(chat_id, content):
    return send_message(chat_id, "chat_id", content)


def _upload_file_http(filename, file_data):
    """通过 HTTP multipart/form-data 直接上传文件（适合所有文件大小）"""
    import urllib.request, urllib.error

    file_size = len(file_data)
    boundary = uuid.uuid4().hex

    # Build multipart form data
    body_parts = []
    body_parts.append(f'--{boundary}\r\nContent-Disposition: form-data; name="file_type"\r\n\r\nstream'.encode())
    body_parts.append(f'--{boundary}\r\nContent-Disposition: form-data; name="file_name"\r\n\r\n{filename}'.encode())
    body_parts.append(f'--{boundary}\r\nContent-Disposition: form-data; name="size"\r\n\r\n{file_size}'.encode())
    header_part = f'--{boundary}\r\nContent-Disposition: form-data; name="file"; filename="{filename}"\r\nContent-Type: application/octet-stream\r\n\r\n'.encode()
    body_parts.append(header_part + file_data + f'\r\n--{boundary}--'.encode())
    body = b'\r\n'.join(body_parts)

    # Get token
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])

    url = 'https://open.feishu.cn/open-apis/im/v1/files?receive_id_type=chat_id'
    req = urllib.request.Request(url, data=body, headers={
        'Authorization': 'Bearer ' + token,
        'Content-Type': f'multipart/form-data; boundary={boundary}'
    })

    try:
        with urllib.request.urlopen(req, timeout=max(60, file_size // 1024 // 100)) as r:
            resp = json.loads(r.read())
            if resp.get('code') == 0:
                return resp['data']['file_key']
            else:
                print(json.dumps({"code": resp.get('code'), "msg": resp.get('msg')}, ensure_ascii=False))
                return None
    except urllib.error.HTTPError as e:
        body_err = e.read().decode('utf-8', errors='replace')
        try:
            err_obj = json.loads(body_err)
            print(json.dumps({"code": err_obj.get('code'), "msg": err_obj.get('msg')}, ensure_ascii=False))
        except Exception:
            print(json.dumps({"code": e.code, "msg": body_err[:500]}, ensure_ascii=False))
        return None


def send_file(chat_id, filepath):
    """上传本地文件并发送到群聊"""
    import io
    if not SDK_AVAILABLE:
        print("错误: 需要安装 lark-oapi SDK，请运行: pip install lark-oapi")
        return

    if not os.path.exists(filepath):
        print(f"错误: 文件不存在: {filepath}")
        return

    filename = os.path.basename(filepath)
    config = load_config()

    client = lark.Client.builder() \
        .app_id(config["app_id"]) \
        .app_secret(config["app_secret"]) \
        .log_level(lark.LogLevel.ERROR) \
        .build()

    # 1. 根据扩展名判断文件类型
    ext = os.path.splitext(filename)[1].lower().lstrip('.')
    image_exts = {'png', 'jpg', 'jpeg', 'gif', 'webp'}

    with open(filepath, 'rb') as f:
        file_data = f.read()

    file_size = len(file_data)
    print(f"文件大小: {file_size/1024/1024:.2f} MB")

    # 图片类型使用 image.create API
    if ext in image_exts:
        from lark_oapi.api.im.v1 import CreateImageRequest, CreateImageRequestBody
        upload_req = CreateImageRequest.builder().request_body(
            CreateImageRequestBody.builder().image_type('message').image(io.BytesIO(file_data)).build()
        ).build()
        upload_resp = client.im.v1.image.create(upload_req)
        if upload_resp.code != 0:
            print(json.dumps({"code": upload_resp.code, "msg": upload_resp.msg}, ensure_ascii=False))
            return
        image_key = upload_resp.data.image_key
        print(f"图片上传成功: {filename} -> image_key: {image_key}")
        content = json.dumps({"image_key": image_key})
        msg_type = "image"
    else:
        # 其他文件使用 HTTP 直传（所有文件均走 im/v1/files 接口）
        file_key = _upload_file_http(filename, file_data)
        if file_key is None:
            return
        print(f"文件上传成功: {filename} -> file_key: {file_key}")
        content = json.dumps({"file_key": file_key})
        msg_type = "file"

    # 2. 发送消息
    from lark_oapi.api.im.v1 import CreateMessageRequest, CreateMessageRequestBody
    msg_request = CreateMessageRequest.builder() \
        .receive_id_type("chat_id") \
        .request_body(CreateMessageRequestBody.builder()
            .receive_id(chat_id)
            .msg_type(msg_type)
            .content(content)
            .build()) \
        .build()

    msg_response = client.im.v1.message.create(msg_request)
    print(json.dumps({"code": msg_response.code, "msg": msg_response.msg}, ensure_ascii=False))


def search_user(keyword):
    import urllib.request
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    encoded_keyword = urllib.parse.quote(keyword)
    url = f"https://open.feishu.cn/open-apis/contact/v3/users/search?query={encoded_keyword}&page_size=20&user_id_type=open_id"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    print(json.dumps(result, ensure_ascii=False, indent=2))
    return result


def search_chat(keyword):
    import urllib.request
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    encoded_keyword = urllib.parse.quote(keyword)
    url = f"https://open.feishu.cn/open-apis/im/v1/chats?search_key={encoded_keyword}&page_size=20"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())
    print(json.dumps(result, ensure_ascii=False, indent=2))
    return result


def get_messages(chat_id, page_size=50):
    """获取群聊消息列表，正确解析并展示中文内容"""
    import urllib.request
    config = load_config()
    token = get_tenant_access_token(config["app_id"], config["app_secret"])
    url = f"https://open.feishu.cn/open-apis/im/v1/messages?container_id_type=chat&container_id={chat_id}&page_size={page_size}"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {token}"})
    with urllib.request.urlopen(req) as resp:
        result = json.loads(resp.read())

    if result.get("code") != 0:
        print(json.dumps(result, ensure_ascii=False, indent=2))
        return result

    items = result.get("data", {}).get("items", [])
    print(f"共获取到 {len(items)} 条消息：")
    print()

    for it in items:
        msg_type = it.get("msg_type")
        sender = it.get("sender", {})
        sender_id = sender.get("id", "")
        sender_type = sender.get("sender_type", "")
        ct = int(it.get("create_time", 0)) // 1000
        dt = datetime.datetime.fromtimestamp(ct).strftime("%Y-%m-%d %H:%M:%S") if ct else "N/A"

        body_str = it.get("body", {}).get("content", "{}")
        body = json.loads(body_str)

        print(f"[{dt}] {sender_type} {sender_id[:16]} | {msg_type}")

        if msg_type == "text":
            text = body.get("text", "")
            print(f"  {text}")
        elif msg_type == "post":
            title = body.get("title", "")
            parts = []
            for section in body.get("content", []):
                for item in section:
                    tag = item.get("tag", "")
                    if tag == "text":
                        parts.append(item.get("text", ""))
                    elif tag == "at":
                        parts.append("@" + item.get("user_id", ""))
                    elif tag == "a":
                        parts.append(item.get("text", "") + "(" + item.get("href", "") + ")")
            content = "".join(parts)
            if title:
                print(f"  Title: {title}")
            print(f"  {content}")
        elif msg_type == "file":
            file_key = body.get("file_key", "")
            print(f"  [文件] key: {file_key}")
        elif msg_type == "system":
            template = body.get("template", "")
            from_user = ",".join(body.get("from_user", []))
            to_chatters = ",".join(body.get("to_chatters", []))
            try:
                content_str = template.format(from_user=from_user, to_chatters=to_chatters)
            except Exception:
                content_str = json.dumps(body, ensure_ascii=False)
            print(f"  {content_str}")
        else:
            print(f"  {json.dumps(body, ensure_ascii=False)[:100]}")
        print()

    return result


if __name__ == "__main__":
    args = sys.argv[1:]
    if len(args) < 1:
        print("用法: python feishu_msg.py <command> [args...]")
        print("可用命令: send-user, send-chat, search-user, search-chat, get-messages, send-file")
        sys.exit(1)

    cmd = args[0]

    if cmd == "send-user" and len(args) >= 3:
        send_user(args[1], args[2])
    elif cmd == "send-chat" and len(args) >= 3:
        send_chat(args[1], args[2])
    elif cmd == "search-user" and len(args) >= 2:
        search_user(args[1])
    elif cmd == "search-chat" and len(args) >= 2:
        search_chat(args[1])
    elif cmd == "get-messages" and len(args) >= 2:
        page_size = int(args[2]) if len(args) >= 3 else 50
        get_messages(args[1], page_size)
    elif cmd == "send-file" and len(args) >= 3:
        send_file(args[1], args[2])
    else:
        print("未知命令或参数不足")
        print("可用命令: send-user, send-chat, search-user, search-chat, get-messages, send-file")
        sys.exit(1)

FILE:scripts/fetch_chat_messages.py
import json, urllib.request, urllib.error, datetime

config = json.load(open('filepath'))
url = 'https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal'
data = json.dumps({'app_id': config['app_id'], 'app_secret': config['app_secret']}).encode()
req = urllib.request.Request(url, data=data, headers={'Content-Type': 'application/json'})
with urllib.request.urlopen(req) as r:
    token = json.loads(r.read())['tenant_access_token']

chat_id = 'oc_3b77de8ba3bc9aa5512abfd8881604bb'
url2 = 'https://open.feishu.cn/open-apis/im/v1/messages?container_id_type=chat&container_id=' + chat_id + '&page_size=50'
req2 = urllib.request.Request(url2, headers={'Authorization': 'Bearer ' + token})
with urllib.request.urlopen(req2) as r:
    result = json.loads(r.read())

items = result.get('data', {}).get('items', [])
print('Total messages:', len(items))
print()

msgs = []
for it in items:
    msg_type = it.get('msg_type')
    sender = it.get('sender', {}).get('id', '')
    ct = int(it.get('create_time', 0)) // 1000
    dt = datetime.datetime.fromtimestamp(ct).strftime('%Y-%m-%d %H:%M:%S') if ct else 'N/A'
    body = json.loads(it.get('body', {}).get('content', '{}'))

    if msg_type == 'text':
        content = body.get('text', '')
        msgs.append({'time': dt, 'type': 'text', 'sender': sender, 'content': content})
        print(f"[text] {dt} | {sender} | {content}")
    elif msg_type == 'post':
        # rich text post
        title = body.get('title', '')
        parts = []
        for section in body.get('content', []):
            for item in section:
                tag = item.get('tag', '')
                if tag == 'text':
                    parts.append(item.get('text', ''))
                elif tag == 'at':
                    parts.append('@' + item.get('user_id', ''))
        content = title + ' ' + ''.join(parts)
        msgs.append({'time': dt, 'type': 'post', 'sender': sender, 'content': content})
        print(f"[post] {dt} | {sender} | {title} | {''.join(parts)}")
    elif msg_type == 'system':
        template = body.get('template', '')
        from_user = ','.join(body.get('from_user', []))
        to_chatters = ','.join(body.get('to_chatters', []))
        content = template.format(from_user=from_user, to_chatters=to_chatters)
        msgs.append({'time': dt, 'type': 'system', 'sender': 'system', 'content': content})
        print(f"[system] {dt} | {content}")

# create document
url3 = 'https://open.feishu.cn/open-apis/docx/v1/documents'
data3 = json.dumps({'title': 'test群聊消息记录'}).encode('utf-8')
req3 = urllib.request.Request(url3, data=data3, method='POST', headers={'Content-Type': 'application/json; charset=utf-8', 'Authorization': 'Bearer ' + token})
with urllib.request.urlopen(req3) as r:
    doc_result = json.loads(r.read())

print()
print('doc create code:', doc_result.get('code'))
if doc_result.get('code') != 0:
    print('failed:', doc_result.get('msg'))
    import sys
    sys.exit(1)

doc_token = doc_result['data']['document']['document_id']
print('doc_token:', doc_token)

# get root block
url4 = 'https://open.feishu.cn/open-apis/docx/v1/documents/' + doc_token + '/blocks?page_size=50'
req4 = urllib.request.Request(url4, headers={'Authorization': 'Bearer ' + token})
with urllib.request.urlopen(req4) as r:
    blocks_result = json.loads(r.read())
root_id = blocks_result['data']['items'][0]['block_id']

def h2(content):
    return {'block_type': 4, 'heading2': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

def bullet(content):
    return {'block_type': 12, 'bullet': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

def text(content):
    return {'block_type': 2, 'text': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

doc_blocks = [h2('群聊消息记录')]
for m in msgs:
    type_icon = {'text': '', 'post': '', 'system': '[系统] '}.get(m['type'], '')
    line = '%(time)s | %(type_icon)s%(content)s' % {'time': m['time'], 'type_icon': type_icon, 'content': m['content']}
    doc_blocks.append(bullet(line))

url5 = 'https://open.feishu.cn/open-apis/docx/v1/documents/' + doc_token + '/blocks/' + root_id + '/children'
data5 = json.dumps({'children': doc_blocks}).encode('utf-8')
req5 = urllib.request.Request(url5, data=data5, method='POST', headers={'Content-Type': 'application/json; charset=utf-8', 'Authorization': 'Bearer ' + token})
with urllib.request.urlopen(req5) as r:
    write_result = json.loads(r.read())

print('write code:', write_result.get('code'))
if write_result.get('code') == 0:
    print('doc_url:', 'https://feishu.cn/docx/' + doc_token)
else:
    print('write failed:', write_result.get('msg'))

FILE:scripts/fix_xlsx_encoding.py
import zipfile, shutil, os, re, tempfile

desktop = os.path.join(os.environ['USERPROFILE'], 'Desktop')
target_size = 15914
matching = [(f, os.path.getsize(os.path.join(desktop, f))) for f in os.listdir(desktop)
            if f.endswith('.xlsx') and os.path.getsize(os.path.join(desktop, f)) == target_size]
filename, _ = matching[0]
filepath = os.path.join(desktop, filename)
print('Fixing:', filepath)

# Read all files from the xlsx
with zipfile.ZipFile(filepath, 'r') as z:
    names = z.namelist()
    files = {}
    for name in names:
        with z.open(name) as f:
            files[name] = f.read()

# Fix sharedStrings.xml: the bytes are GBK but XML declares UTF-8
# Re-decode as GBK and re-encode as UTF-8
shared = files['xl/sharedStrings.xml']
# Decode the raw bytes as GBK, then encode as UTF-8
content_str = shared.decode('gbk', errors='replace')
# Remove the encoding declaration and save as UTF-8
content_str = content_str.replace('encoding="UTF-8"', 'encoding="UTF-8"')
# Re-encode only the content (t tags) as proper UTF-8
def fix_t(content):
    # For each <t> element, get the text, decode from GBK bytes (the bytes we have are GBK), 
    # then the string is already decoded - but we need to re-encode properly
    def replacer(m):
        return m.group(0)
    return content

# Actually, let's take a different approach: rebuild the strings correctly
# The sharedStrings.xml stores raw UTF-8 bytes but labeled as such - the issue is the underlying storage
# Let's just re-save the strings by reading them correctly (as GBK) and writing as UTF-8

new_shared = shared.decode('gbk', errors='replace').encode('utf-8', errors='replace')
files['xl/sharedStrings.xml'] = new_shared

# Also fix worksheet and other XMLs that may have inline strings
for name in ['xl/worksheets/sheet1.xml', 'xl/workbook.xml']:
    if name in files:
        content = files[name]
        try:
            s = content.decode('gbk', errors='replace')
            files[name] = s.encode('utf-8', errors='replace')
        except:
            pass

# Save fixed xlsx
fixed_path = os.path.join(desktop, 'fixed_' + filename)
with zipfile.ZipFile(fixed_path, 'w', zipfile.ZIP_DEFLATED) as zout:
    for name in names:
        zout.writestr(name, files[name])

print('Fixed file saved to:', fixed_path)
print('Original size:', os.path.getsize(filepath))
print('Fixed size:', os.path.getsize(fixed_path))

FILE:scripts/rebuild_xlsx.py
import openpyxl
from openpyxl.styles import Font, Alignment, PatternFill
import datetime, os, json, lark_oapi as lark
from lark_oapi.api.im.v1 import *

# Read original file
filepath = 'C:/Users/10430/Desktop/采购数据.xlsx'
wb_src = openpyxl.load_workbook(filepath)
ws_src = wb_src.active
rows = list(ws_src.iter_rows(values_only=True))
print(f'Read {len(rows)} rows from original file')

# Create new clean xlsx
wb_new = openpyxl.Workbook()
ws_new = wb_new.active
ws_new.title = '采购数据'

# Styles
header_font = Font(bold=True)
header_fill = PatternFill(start_color='D9E1F2', end_color='D9E1F2', fill_type='solid')
center = Alignment(horizontal='center', vertical='center')

# Write header
headers = ['采购时间', '部门', '项目', '采购人', '花费']
for col, h in enumerate(headers, 1):
    cell = ws_new.cell(row=1, column=col, value=h)
    cell.font = header_font
    cell.fill = header_fill
    cell.alignment = center

# Write data rows
for r_idx, row in enumerate(rows[1:], 2):  # skip header row
    for c_idx, val in enumerate(row, 1):
        if isinstance(val, datetime.datetime):
            ws_new.cell(row=r_idx, column=c_idx, value=val).number_format = 'YYYY-MM-DD'
        else:
            ws_new.cell(row=r_idx, column=c_idx, value=val)

# Auto-adjust column widths
for col in ws_new.columns:
    max_len = 0
    col_letter = col[0].column_letter
    for cell in col:
        if cell.value:
            max_len = max(max_len, len(str(cell.value)))
    ws_new.column_dimensions[col_letter].width = min(max_len + 4, 30)

# Save locally
out_path = 'C:/Users/10430/Desktop/采购数据_干净版.xlsx'
wb_new.save(out_path)
print(f'Saved to: {out_path}')

# Upload via lark-oapi SDK
config = json.load(open('C:/Users/10430/.openclaw/workspace/skills/feishu-ops/scripts/config.json', encoding='utf-8'))
client = lark.Client.builder() \
    .app_id(config['app_id']) \
    .app_secret(config['app_secret']) \
    .log_level(lark.LogLevel.ERROR) \
    .build()

with open(out_path, 'rb') as f:
    file_data = f.read()

upload_resp = client.im.v1.file.create(
    CreateFileRequest.builder()
        .request_body(CreateFileRequestBody.builder()
            .file_type('xlsx')
            .file_name('采购数据_干净版.xlsx')
            .file(file_data)
            .build())
        .build()
)
print('Upload code:', upload_resp.code)
if upload_resp.code != 0:
    print('Upload failed:', upload_resp.msg)
    import sys
    sys.exit(1)

file_key = upload_resp.data.file_key
print('file_key:', file_key)

# Send to chat
msg_resp = client.im.v1.message.create(
    CreateMessageRequest.builder()
        .receive_id_type('chat_id')
        .request_body(CreateMessageRequestBody.builder()
            .receive_id('oc_2c6df8f6e06e88d34729baacc124b89e')
            .msg_type('file')
            .content(json.dumps({'file_key': file_key}))
            .build())
        .build()
)
print('Send code:', msg_resp.code, 'msg:', msg_resp.msg)
print('Done!')

FILE:scripts/send_news.py
import json, urllib.request, uuid

config = json.load(open('C:/Users/10430/.openclaw/workspace/skills/feishu-ops/scripts/config.json', encoding='utf-8'))
url = 'https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal'
data = json.dumps({'app_id': config['app_id'], 'app_secret': config['app_secret']}).encode('utf-8')
req = urllib.request.Request(url, data=data, headers={'Content-Type': 'application/json'})
with urllib.request.urlopen(req) as r:
    token = json.loads(r.read())['tenant_access_token']

news = """以下是今日国际热点新闻 Top 10（2026年3月19日）：

1. 中美巴黎经贸磋商释放积极信号，美方希望双边关系稳定发展
2. 以军称伊朗使用集束弹头导弹再次发动袭击，中东局势持续升级
3. 美情报总监称伊朗政权"完整但遭严重削弱"
4. 美军多处基地遭袭击，"被炸了个遍"
5. 世卫组织正为中东地区可能发生的"核灾难"做准备
6. 日本央行维持政策利率在0.75%左右
7. 卡塔尔多处能源设施再遭袭起火，阿联酋谴责
8. 一艘船在阿联酋海域遭袭起火，波斯湾局势紧张
9. 美联储按兵不动，美元大涨，人民币对美元汇率跌破6.9
10. 高市启"跪美"外交，引发国际关注

—— 由AI龙虾自动整理推送"""

url2 = 'https://open.feishu.cn/open-apis/im/v1/messages?receive_id_type=chat_id'
payload = {
    'receive_id': 'oc_2c6df8f6e06e88d34729baacc124b89e',
    'msg_type': 'text',
    'content': json.dumps({'text': news}),
    'uuid': str(uuid.uuid4())
}
data2 = json.dumps(payload).encode('utf-8')
req2 = urllib.request.Request(url2, data=data2, method='POST', headers={'Content-Type': 'application/json; charset=utf-8', 'Authorization': 'Bearer ' + token})
with urllib.request.urlopen(req2) as r:
    result = json.loads(r.read())
print('code:', result.get('code'), '| msg:', result.get('msg'))

FILE:scripts/test_blocks.py
import json, urllib.request, urllib.error

config = json.load(open('C:/Users/10430/.openclaw/workspace/skills/feishu-ops/scripts/config.json'))
token = 't-g1043jexJNXU6IOGYC26VWSWESLRGIMMOWRXJWYS'
doc_token = 'KH6qdDlNuoACh4xSaPTcFFr3npc'

tests = [
    (3, 'heading1'),
    (4, 'heading2'),
    (5, 'heading3'),
    (6, 'bullet'),
    (7, 'ordered'),
]

for bt, key in tests:
    blocks = [{'block_type': bt, key: {'elements': [{'type': 'text_run', 'text_run': {'content': 'Test content'}}], 'style': {}}}]
    url = 'https://open.feishu.cn/open-apis/docx/v1/documents/' + doc_token + '/blocks/' + doc_token + '/children'
    data = json.dumps({'children': blocks}).encode('utf-8')
    req = urllib.request.Request(url, data=data, method='POST', headers={'Content-Type': 'application/json', 'Authorization': 'Bearer ' + token})
    try:
        with urllib.request.urlopen(req) as r:
            result = json.loads(r.read())
        code = result.get('code')
        print('block_type', bt, '(' + key + '): OK code=' + str(code))
    except urllib.error.HTTPError as e:
        body = e.read().decode()
        print('block_type', bt, '(' + key + '): Error ' + str(e.code) + ' - ' + body[:100])

FILE:scripts/test_blocks2.py
import json, urllib.request, urllib.error

config = json.load(open('C:/Users/10430/.openclaw/workspace/skills/feishu-ops/scripts/config.json'))
token = 't-g1043jexJNXU6IOGYC26VWSWESLRGIMMOWRXJWYS'
doc_token = 'KH6qdDlNuoACh4xSaPTcFFr3npc'

url = 'https://open.feishu.cn/open-apis/docx/v1/documents/' + doc_token + '/blocks/' + doc_token + '/children'

tests = [
    ({'block_type': 6, 'bullet': {'elements': [{'type': 'text', 'text': {'content': 'Bullet item'}}], 'style': {}}}, 'bullet with text type'),
    ({'block_type': 9, 'code': {'elements': [{'type': 'text_run', 'text_run': {'content': 'Code line'}}], 'style': {'language': 1}}}, 'code block'),
    ({'block_type': 10, 'quote': {'elements': [{'type': 'text_run', 'text_run': {'content': 'Quote text'}}], 'style': {}}}, 'quote block'),
    ({'block_type': 2, 'paragraph': {'elements': [{'type': 'text_run', 'text_run': {'content': 'Para text'}}], 'style': {}}}, 'paragraph'),
]

for block, desc in tests:
    data = json.dumps({'children': [block]}).encode('utf-8')
    req = urllib.request.Request(url, data=data, method='POST', headers={'Content-Type': 'application/json', 'Authorization': 'Bearer ' + token})
    try:
        with urllib.request.urlopen(req) as r:
            result = json.loads(r.read())
        print(desc + ': OK code=' + str(result.get('code')))
    except urllib.error.HTTPError as e:
        body = e.read().decode()
        print(desc + ': Error ' + str(e.code) + ' - ' + body[:120])

FILE:scripts/upload_sdk.py
import lark_oapi as lark
from lark_oapi.api.im.v1 import *
import json, os

config = json.load(open('C:/Users/10430/.openclaw/workspace/skills/feishu-ops/scripts/config.json', encoding='utf-8'))

client = lark.Client.builder() \
    .app_id(config['app_id']) \
    .app_secret(config['app_secret']) \
    .log_level(lark.LogLevel.ERROR) \
    .build()

desktop = os.path.join(os.environ['USERPROFILE'], 'Desktop')
target_size = 15914
matching = [(f, os.path.getsize(os.path.join(desktop, f))) for f in os.listdir(desktop) 
            if f.endswith('.xlsx') and os.path.getsize(os.path.join(desktop, f)) == target_size]
filename, _ = matching[0]
filepath = os.path.join(desktop, filename)
print('File:', filepath)

# Upload file using SDK
with open(filepath, 'rb') as f:
    file_data = f.read()

request = CreateFileRequest.builder() \
    .request_body(CreateFileRequestBody.builder()
        .file_type("xlsx")
        .file_name(filename)
        .file(file_data)
        .build()) \
    .build()

response = client.im.v1.file.create(request)
print('code:', response.code, 'msg:', response.msg)
if response.success():
    print('file_key:', response.data.file_key)
else:
    print('raw:', response.raw.content[:200] if hasattr(response, 'raw') else 'no raw')

FILE:scripts/upload_test.py
import json, os, requests, urllib.parse

config = json.load(open('C:/Users/10430/.openclaw/workspace/skills/feishu-ops/scripts/config.json', encoding='utf-8'))
r = requests.post('https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal', 
                  json={'app_id': config['app_id'], 'app_secret': config['app_secret']})
token = r.json()['tenant_access_token']

desktop = os.path.join(os.environ['USERPROFILE'], 'Desktop')
target_size = 15914
matching = [(f, os.path.getsize(os.path.join(desktop, f))) for f in os.listdir(desktop) 
            if f.endswith('.xlsx') and os.path.getsize(os.path.join(desktop, f)) == target_size]
filename, _ = matching[0]
filepath = os.path.join(desktop, filename)
encoded_name = urllib.parse.quote(filename)

# Use data= instead of files= for file upload (plain binary)
url = f'https://open.feishu.cn/open-apis/im/v1/files?file_name={encoded_name}&file_type=xlsx'
with open(filepath, 'rb') as f:
    file_data = f.read()

headers = {
    'Authorization': 'Bearer ' + token,
    'Content-Type': 'application/octet-stream'
}
r2 = requests.post(url, data=file_data, headers=headers)
print('data=bytes:', r2.status_code, r2.json().get('code'), r2.json().get('msg', '')[:80])

# Try multipart with correct format
url3 = f'https://open.feishu.cn/open-apis/im/v1/files?file_name={encoded_name}&file_type=xlsx'
with open(filepath, 'rb') as f:
    files = {'file': (filename, f)}
    r3 = requests.post(url3, files=files, headers={'Authorization': 'Bearer ' + token})
print('multipart:', r3.status_code, r3.json().get('code'), r3.json().get('msg', '')[:80])

# Try with explicit filename in multipart
url4 = f'https://open.feishu.cn/open-apis/im/v1/files?file_name={encoded_name}&file_type=xlsx'
with open(filepath, 'rb') as f:
    files = {'file': (filename, f, 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet')}
    r4 = requests.post(url4, files=files, headers={'Authorization': 'Bearer ' + token})
print('multipart+ct:', r4.status_code, r4.json().get('code'), r4.json().get('msg', '')[:80])

FILE:scripts/write_news.py
import json, urllib.request, urllib.error

config = json.load(open('C:/Users/10430/.openclaw/workspace/skills/feishu-ops/scripts/config.json'))
token = 't-g1043jexJNXU6IOGYC26VWSWESLRGIMMOWRXJWYS'
doc_token = 'KH6qdDlNuoACh4xSaPTcFFr3npc'

# get root block id
url = 'https://open.feishu.cn/open-apis/docx/v1/documents/' + doc_token + '/blocks?page_size=50'
req = urllib.request.Request(url, headers={'Authorization': 'Bearer ' + token})
with urllib.request.urlopen(req) as r:
    result = json.loads(r.read())

root_id = result['data']['items'][0]['block_id']
print('root block:', root_id)

def h1(content):
    return {'block_type': 3, 'heading1': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

def h2(content):
    return {'block_type': 4, 'heading2': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

def bullet(content):
    return {'block_type': 12, 'bullet': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

def ordered(content):
    return {'block_type': 13, 'ordered': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

def text(content):
    return {'block_type': 2, 'text': {'elements': [{'type': 'text_run', 'text_run': {'content': content}}]}}

news = [
    h1('今日热点新闻汇总'),
    h2('国内要闻'),
    bullet('一批重大工程紧锣密鼓，奏响"春日奋进曲"'),
    bullet('王毅表示：这场战争本不该发生，更没必要再打下去'),
    bullet('"投资者保护"首次写入十五五，证监会火速部署清除"拦路虎"'),
    bullet('商办最低首付调至3成，北上广杭蓉锁定一线城市'),
    bullet('多地宣布27年缩减中考计分科目'),
    h2('国际热点'),
    bullet('伊朗外长怒斥以色列"暗杀常态化"，称国际社会不应"双标"'),
    bullet('美媒：世卫组织官员正为中东地区可能发生的"核灾难"做准备'),
    bullet('牛弹琴：重大转折点，战争进入疯狂阶段'),
    bullet('中国代表：中东局势正被推向危险深渊'),
    h2('科技财经'),
    bullet('Agent推高需求，全球云计算巨头集体涨价'),
    bullet('腾讯2025年营收增长14%，今年AI新产品投入至少翻倍'),
    bullet('苹果CEO库克现身成都，锚定中国供应链'),
    bullet('金价大跌，市民排队买金（现货黄金跌破4900美元关口）'),
    h2('社会民生'),
    bullet('前中超球员邱忠辉：送外卖不丢人，凭劳动挣钱最光荣'),
    bullet('央媒：是时候打破眼镜行业的价格"滤镜"了'),
    bullet('胖东来169元1克拉方糖戒指再上架，每人限购5枚'),
    h2('体育速递'),
    bullet('唐钱婷刷新50蛙亚洲纪录（预赛29秒49）'),
    bullet('东契奇&库里膝盖出血：地板擦伤留下的真实血迹'),
    text('—— 以上新闻整理自新浪网，采集时间：2026年3月19日'),
]

url2 = 'https://open.feishu.cn/open-apis/docx/v1/documents/' + doc_token + '/blocks/' + root_id + '/children'
data = json.dumps({'children': news}).encode('utf-8')
req2 = urllib.request.Request(url2, data=data, method='POST', headers={'Content-Type': 'application/json; charset=utf-8', 'Authorization': 'Bearer ' + token})
try:
    with urllib.request.urlopen(req2) as r:
        result2 = json.loads(r.read())
    code = result2.get('code')
    children = result2.get('data', {}).get('children', [])
    print('code:', code, '| children count:', len(children))
    if code != 0:
        print('msg:', result2.get('msg'))
except urllib.error.HTTPError as e:
    body = e.read().decode()
    print('HTTP Error:', e.code)
    print('Body:', body[:500])

FILE:references/api_ref.md
# 飞书 API 参考

## 认证

```
POST https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal
Body: {"app_id": "...", "app_secret": "..."}
返回: {"code": 0, "tenant_access_token": "..."}
```

## 文档 API

### 创建文档
```
POST https://open.feishu.cn/open-apis/docx/v1/documents
Headers: Authorization: Bearer <token>
Body: {"title": "文档标题"}
返回: {"code": 0, "data": {"document": {"document_id": "xxx"}}}
```

### 读取文档 blocks
```
GET https://open.feishu.cn/open-apis/docx/v1/documents/<doc_token>/blocks?page_size=500
返回: {"code": 0, "data": {"items": [block, ...]}}
```

### 查询单个 block
```
GET https://open.feishu.cn/open-apis/docx/v1/documents/<doc_token>/blocks/<block_id>
```

### 追加 blocks（到父 block）
```
POST https://open.feishu.cn/open-apis/docx/v1/documents/<doc_token>/blocks/<parent_block_id>/children
Body: {"children": [block, ...], "index": 0}
```

### 更新 block 文本
```
PATCH https://open.feishu.cn/open-apis/docx/v1/documents/<doc_token>/blocks/<block_id>
Body: {"update_text_elements": {"elements": [...]}}
```

### 删除 block
```
DELETE https://open.feishu.cn/open-apis/docx/v1/documents/<doc_token>/blocks/<block_id>
```

## 消息 API

### 发送消息
```
POST https://open.feishu.cn/open-apis/im/v1/messages?receive_id_type=<type>
Headers: Authorization: Bearer <token>
Body: {
  "receive_id": "<id>",
  "msg_type": "text",
  "content": "{\"text\": \"消息内容\"}",
  "uuid": "<uuid>"
}
```
- `receive_id_type`: `open_id`（用户）或 `chat_id`（群）

### 搜索用户
```
GET https://open.feishu.cn/open-apis/contact/v3/users/search?query=<关键词>&page_size=20
```

### 搜索群聊
```
GET https://open.feishu.cn/open-apis/im/v1/chats?search_key=<关键词>&page_size=20
```

## Block 类型参考

| block_type | 类型 | 键名 |
|-----------|------|------|
| 1 | Page | `page` |
| 2 | Text | `text` |
| 3 | Heading 1 | `heading1` |
| 4 | Heading 2 | `heading2` |
| 5 | Heading 3 | `heading3` |
| 6-11 | Heading 4-9 | `heading4`...`heading9` |
| 12 | Bullet List | `bullet` |
| 13 | Ordered List | `ordered` |
| 14 | Code | `code` |
| 15 | Quote | `quote` |
| 17 | Todo | `todo` |
| 31 | Table | `table` |
| 32 | Table Cell | `table_cell` |

## 表格 block 内部结构

表格 block (type=11) 包含 `table_rows`，每个 row 包含 `cells`。

查询表格内容需要：
1. `GET /blocks` 获取文档 blocks
2. 找到 `block_type == 11` 的表格 block
3. `GET /blocks/<table_block_id>` 获取表格详情（含 rows/cells）
4. 每个 cell 是独立的 block，可用 `update_text_elements` 更新

## 错误码

| code | 含义 |
|------|------|
| 0 | 成功 |
| 99991661 | 无权限 |
| 230001 | 文档不存在 |
| 230002 | block 不存在 |