小红书

Xiaohongshu Skill

小红书（XiaoHongShu / Little Red Book）数据采集和交互工具包。基于RedCrack纯Python逆向工程实现。

Quick Start

Installation

Dependencies are already installed:

pip install aiohttp loguru pycryptodome getuseragent

Basic Usage

import asyncio
import sys
sys.path.insert(0, r'C:\\Users\\Chocomint\\.openclaw\\workspace\\xiaohongshu\\scripts')

from request.web.xhs_session import create_xhs_session

async def main():
    # ✅ 推荐：不强制代理（有代理再填 proxy）
    # 说明：当前小红书接口经常对“未登录/游客”限制搜索能力。
    # 如果 search 报 code=-104（未登录无权限），请提供 web_session。
    xhs = await create_xhs_session(proxy=None, web_session="YOUR_WEB_SESSION_OR_NONE")

    # Search notes
    res = await xhs.apis.note.search_notes("美妆")
    data = await res.json()
    print(data)

    await xhs.close_session()

asyncio.run(main())

Core Capabilities

1. Search & Discovery

Search notes by keyword:

res = await xhs_session.apis.note.search_notes("口红")

Get home feed (trending):

注意：get_homefeed 需要 category 参数
res = await xhs_session.apis.note.get_homefeed(
    xhs_session.apis.note.homefeed_category_enum.FOOD
)

Get note detail:

note_detail 需要 note_id + xsec_token（有时在搜索结果 item 里叫 xsec_token）
res = await xhs_session.apis.note.note_detail(note_id, xsec_token)

2. User Interactions

Get user info:

res = await xhs_session.apis.auth.get_self_simple_info()

Follow a user:

res = await xhs_session.apis.user.follow_user(user_id)

Like a note:

res = await xhs_session.apis.note.like_note(note_id)

3. Comments

Get comments for a note:

comments 也需要 note_id + xsec_token
res = await xhs_session.apis.comments.get_comments(note_id, xsec_token)

Configuration

Proxy

代理不是硬性要求（本技能可以 proxy=None 运行）。但在以下情况建议使用代理：

• -网络环境不稳定/请求超时
• -频繁触发风控（例如 461）想换出口 IP 试试

不使用代理：

xhs = await create_xhs_session(proxy=None, web_session=None)

使用代理：

xhs = await create_xhs_session(
    proxy="http://127.0.0.1:7890",
    web_session="your_web_session_cookie"  # 需要登录能力时再提供
)

Encryption Parameters

All encryption parameters are automatically generated:

• -Cookies: a1, webId, acw_tc, web_session, sec_poison_id, websectiga, gid
• -Headers: x-s, x-s-common, x-t, x-b3-traceid, x-xray-traceid

Configuration file: scripts/request/web/encrypt/web_encrypt_config.ini

Session Management

Guest Mode (No Login)

xhs_session = await create_xhs_session(proxy="http://127.0.0.1:7890")

Authenticated Mode (With Login)

xhs_session = await create_xhs_session(
    proxy="http://127.0.0.1:7890",
    web_session="030037afxxxxxxxxxxxxxxxxxxxaeb59d5b4"  # Your cookie
)

Close Session

await xhs_session.close_session()

Links & IDs (重要)

• -小红书笔记的公开标识通常是 note_id（十六进制风格字符串），例如：697cc945000000000a02cdad。
• -这个 note_id 可以做数学意义的 16 进制→10 进制转换，但那只是“另一种表示法”，不会变成小红书的短数字 ID（类似 App Store 的 id741292507），也不会更适合拼接小红书链接。
• -我们通过接口得到的可直接打开的网页链接通常形如：

- https://www.xiaohongshu.com/explore/<note_id>?xsec_token=...&xsec_source=pc_search

• -xhslink.com 短链一般需要在 App/登录态里通过“分享→复制链接”获得；仅靠当前接口通常拿不到。

输出到聊天时的链接美化（推荐）

为了避免长链接难看，优先用文本标签超链接：

• -Markdown：[标题](https://www.xiaohongshu.com/explore/...)

---

Available APIs

All APIs are accessible via xhs_session.apis.*:

Authentication (apis.auth):

• -get_self_simple_info() - Get current user info

Notes (apis.note):

• -search_notes(keyword) - Search notes by keyword
• -get_homefeed(category) - Get home feed
• -note_detail(note_id, share_token) - Get note details
• -like_note(note_id) - Like a note

Comments (apis.comments):

• -get_comments(note_id, share_token) - Get note comments

User (apis.user):

• -follow_user(user_id) - Follow a user
• -get_user_info(user_id) - Get user details

Example Workflows

Workflow 1: Search and Extract Notes

async def search_example():
    xhs_session = await create_xhs_session(proxy="http://127.0.0.1:7890")

    # Search for makeup tutorials
    res = await xhs_session.apis.note.search_notes("美妆教程")
    data = await res.json()

    for note in data['data']['items']:
        print(f"Title: {note['display_title']}")
        print(f"Author: {note['user']['nickname']}")
        print(f"Likes: {note['liked_count']}")
        print("---")

    await xhs_session.close_session()

Workflow 2: Get Comments for Analysis

async def comments_example():
    xhs_session = await create_xhs_session(proxy="http://127.0.0.1:7890")

    note_id = "64f1a2d30000000013003689"
    res = await xhs_session.apis.comments.get_comments(note_id, "")
    data = await res.json()

    for comment in data['data']['comments']:
        print(f"User: {comment['user']['nickname']}")
        print(f"Content: {comment['content']}")
        print(f"Likes: {comment['like_count']}")
        print("---")

    await xhs_session.close_session()

Workflow 3: User Profile Analysis

async def profile_example():
    xhs_session = await create_xhs_session(
        proxy="http://127.0.0.1:7890",
        web_session="your_cookie_here"
    )

    # Get self info
    res = await xhs_session.apis.auth.get_self_simple_info()
    data = await res.json()

    print(f"Username: {data['data']['user']['nickname']}")
    print(f"Followers: {data['data']['user']['follows']}")
    print(f"Fans: {data['data']['user']['fans']}")

    await xhs_session.close_session()

Important Notes

1. Proxy is required for most operations due to XiaoHongShu's anti-scraping measures

2. Rate limiting: Be respectful with request frequency to avoid IP bans

3. Authentication: Some operations require login (web_session cookie)

4. Legal compliance: Use only for legitimate research and data analysis purposes

Technical Details

Based on [RedCrack](https://github.com/Cialle/RedCrack) - Pure Python reverse engineering of XiaoHongShu's encryption algorithms.

What's automatically handled:

• -Base64/Base58 custom encoding
• -RC4/XOR encryption
• -MD5/SHA256 hashing
• -Custom signature generation (x-s, x-s-common)
• -Dynamic cookie generation (a1, webId, sec_poison_id, etc.)

No JavaScript runtime required - All encryption is pure Python.

Troubleshooting

Connection errors

• -Verify your proxy is running on the configured port
• -Try different proxy servers if needed
• -Check network connectivity

461 errors（风控/安全校验）

• -这通常不是代码语法问题，而是触发了小红书的风控/安全校验。
• -典型现象：OtherStatusError: 461异常，或者接口返回看似 success=true 但 HTTP=461。

应对建议：

• -降低频率/加随机 sleep、避免并发
• -换关键词/换 endpoint（例如先用搜索拿到 note_id + xsec_token，再查 detail）
• -使用稳定的登录态（web_session）
• -必要时更换代理出口

401/403 errors

• -web_session 可能过期
• -小红书可能更新了风控参数/签名逻辑（需要更新逆向实现）

Import errors

• -Ensure all dependencies are installed: pip install aiohttp loguru pycryptodome getuseragent
• -Check that the skill path is correct in sys.path.insert()