\r \r

XingTu Task Author Fetcher\r

\r Fetch the complete list of registered authors/influencers for a XingTu (星图) recruitment task and export to a well-formatted Excel file.\r \r

Workflow\r

\r

Step 1: Obtain Task ID\r

\r If the user has already provided a XingTu task ID in their message, use it directly. Otherwise, ask the user to provide the task ID.\r \r The task ID is a long numeric string (e.g., 7642279680695484426). If the user provides a URL or screenshot, extract the ID from it.\r \r

Step 2: Check Cookie\r

\r The cookie file is stored at ~/.xingtuCookie.txt. Use os.path.expanduser("~") to resolve the home directory on any OS.\r \r If the file exists:\r

• Read the cookie content.\r
• Validate the cookie by running:\r

  python "\x3Cskill-base>/scripts/fetch_xingtu_authors.py" --task-id "\x3Ctask_id>" --cookie "\x3Ccookie>" --validate\r
  ```\r
  

• If the script exits with code 0 (prints [OK]), the cookie is valid. Proceed to Step 4.\r
• If the script exits with code 1 (prints [FAIL]), the cookie is invalid. Go to Step 3 with the message: "星图后台登录失效"\r \r If the file does NOT exist:\r
• Go to Step 3 with the message: "星图后台还未登录"\r \r

Step 3: Login Flow\r

\r When the cookie is missing or invalid:\r \r

1. First, check if the user's current message contains a cookie string. If yes:\r
   • Validate it using the --validate command from Step 2.\r
   • If valid: write it to ~/.xingtuCookie.txt and proceed to Step 4.\r
   • If invalid: continue below.\r \r
2. Tell the user the login is required. Use the appropriate message:\r
   • No cookie file: "星图后台还未登录，需要先登录星图后台。"\r
   • Invalid cookie: "星图后台登录失效，需要重新登录。"\r \r
3. Open the login page in the built-in browser:\r

   https://sso.oceanengine.com/xingtu/login?role=7\r
   ```\r
   Use the `agent-browser` skill or `preview_url` tool to open this URL.\r
   

\r 4. Instruct the user:\r

• "请在打开的浏览器中完成星图后台登录。"\r
• "登录成功后，请从浏览器中复制完整的 Cookie 字符串发给我。"\r
• Provide guidance: In Chrome DevTools, go to Application > Cookies > www.xingtu.cn, or use the Network tab to copy the Cookie header from any API request.\r \r

5. When the user provides the cookie string:\r
   • Save it to ~/.xingtuCookie.txt\r
   • Validate it using the --validate command\r
   • If valid: proceed to Step 4\r
   • If still invalid: ask the user to double-check the cookie and try again\r \r

Step 4: Fetch Authors\r

\r Run the fetch script to get all authors across all pages:\r \r

python "\x3Cskill-base>/scripts/fetch_xingtu_authors.py" --task-id "\x3Ctask_id>" --cookie "\x3Ccookie>"\r
```\r
\r
The script will:\r
- Paginate through all results automatically (page by page until `has_more` is false)\r
- Print progress for each page (e.g., "Got 10 authors, cumulative: 10/45")\r
- Export all data to a formatted Excel file\r
- Print a JSON summary with `task_id`, `total_authors`, and `output` path\r
\r
The default output path is: `xingtu_authors_\x3Ctask_id>_\x3Ctimestamp>.xlsx` in the current working directory.\r
\r
Optional: Use `--output \x3Cpath>` to specify a custom output path.\r
\r
### Step 5: Present Results\r
\r
After the script completes successfully:\r
\r
1. Show the summary to the user (total authors count)\r
2. Call `deliver_attachments` to deliver the Excel file\r
3. Call `open_result_view` to open the Excel file for the user\r
\r
### Step 6: Memory Update\r
\r
Append a note to the daily memory file recording:\r
- Task ID processed\r
- Number of authors fetched\r
- Output file path\r
\r
## Excel Output Columns\r
\r
| Column | Source Field |\r
|--------|-------------|\r
| 序号 | Auto-increment |\r
| 达人昵称 | `author_base_info.nick_name` |\r
| 作者ID | `author_base_info.author_id` |\r
| 达人等级 | `author_base_info.ecom_author_level` |\r
| 粉丝数 | `author_base_info.follower` |\r
| 主推类目(30天) | `author_base_info.all_ecom_top3_category_30d_desc` (joined) |\r
| 带货GMV(30天) | `author_base_info.all_ecom_gmv_30d_desc` |\r
| 视频GMV(30天) | `author_base_info.ecom_video_gmv_30d_desc` |\r
| 1-20s报价 | `recruit_author_order_info.recruit_cpt_info.author_price` |\r
| 21-60s报价 | `author_base_info.price_21_60` |\r
| 60s+报价 | `author_base_info.price_60` |\r
| 预期CPM | `author_base_info.prospective_cpm` |\r
| 预期播放量 | `author_base_info.expected_play_num` |\r
| 完播率 | `author_base_info.author_recruit_video_cpt_fulfillment_rate_desc` |\r
| 所在城市 | `author_base_info.author_resident_city` |\r
| 内容标签 | `author_base_info.content_tags` (joined) |\r
| 微信号 | `author_base_info.wechat` |\r
| 报名状态 | `recruit_author_order_info.enroll_status` (mapped) |\r
| 报名时间 | `recruit_author_order_info.time_info.enroll_time` (timestamp to datetime) |\r
| 推荐理由 | `recruit_author_order_info.recommend_reason` (formatted) |\r
| 合作优势 | `recruit_author_order_info.coop_advantage` |\r
| 创作思路 | `recruit_author_order_info.creation_idea` |\r
| 补充说明 | `recruit_author_order_info.extra_note` |\r
\r
## API Reference\r
\r
- **Endpoint**: `POST https://www.xingtu.cn/gw/api/challenge/provider_get_task_author_list`\r
- **Auth**: Cookie-based (from browser login at `https://sso.oceanengine.com/xingtu/login?role=7`)\r
- **Pagination**: Query parameter `page` (int, starts at 1), response field `pagination.has_more` (bool)\r
- **Required headers**: `Accept`, `Content-Type`, `agw-js-conv: str`, `Cookie`, `User-Agent`, `Host`\r
\r
## Error Handling\r
\r
- **Cookie expired mid-fetch**: If any page returns a non-zero `status_code`, report the error and suggest re-login\r
- **Empty result**: If `total_count` is 0, inform the user that no authors have registered for this task\r
- **Network errors**: Retry once; if still failing, report the error to the user\r
- **openpyxl not installed**: Install it with `pip install openpyxl requests`\r