← 返回 Skills 市场
Xiaohongshu Post (Browser Auto)
作者
antaeus001
· GitHub ↗
· v1.0.7
· MIT-0
259
总下载
0
收藏
2
当前安装
8
版本数
在 OpenClaw 中安装
/install xiaohongshu-post-antaeus
功能描述
Create and publish Xiaohongshu (小红书/RED) notes. Use when: user wants to publish a note to Xiaohongshu, schedule or batch post content, generate Xiaohongshu-s...
安全使用建议
This skill is coherent with its stated purpose, but take care with browser automation: it reads page HTML (which can include login tokens or sensitive data) and will send sanitized HTML to an LLM endpoint if you configure DASHSCOPE_API_KEY/OPENAI_API_KEY or a remote XHS_ANALYZER_BASE_URL. To minimize risk, prefer API mode (set XIAOHONGSHU_APP_KEY/SECRET), or run the analyzer locally (Ollama) by setting XHS_ANALYZER_BASE_URL to a localhost endpoint and not providing external API keys. Review the code if you plan to supply any analyzer API keys, run in an isolated environment, and consider using --draft-only for manual posting if you do not trust external LLM endpoints.
功能分析
Type: OpenClaw Skill
Name: xiaohongshu-post-antaeus
Version: 1.0.7
The skill bundle provides legitimate tools for automating Xiaohongshu (RED) posts via official APIs or browser automation using Playwright. While the browser mode involves sending page HTML to an LLM for analysis (page_analyzer.py), the developers included explicit security warnings in SKILL.md and implemented a sanitization function (_sanitize_html) to strip sensitive tokens and cookies before transmission. The code is well-structured, lacks obfuscation, and focuses entirely on the stated purpose of content publishing.
能力评估
Purpose & Capability
Name/description match the code and instructions: post.py implements official-platform API posting and draft mode; publish_browser.py implements Playwright-driven automation with an LLM-based page analyzer. Required binary (python3) and listed optional env vars (platform credentials, analyzer keys, proxies) are appropriate for these modes.
Instruction Scope
Browser mode explicitly grabs page HTML and sends a sanitized version to an LLM endpoint (DashScope/OpenAI or a configured local analyzer). The SKILL.md warns about sensitive data in pages and offers a local-Ollama fallback; this behavior is necessary for the model-driven automation but creates a privacy/exfiltration risk if you provide an external API key. The code performs HTML sanitization (removes script/style and token-like patterns) but notes it cannot guarantee full removal of sensitive info.
Install Mechanism
There is no formal install spec in registry metadata, but requirements.txt (playwright, openai) and SKILL.md instruct pip install and 'playwright install chromium'. This is a standard dependency flow; Playwright will download browser binaries (Chromium). No downloads from arbitrary/personal URLs are present in the repo.
Credentials
Environment variables are optional and consistent with functionality: XIAOHONGSHU_APP_KEY/SECRET for official API; DASHSCOPE_API_KEY / OPENAI_API_KEY / XHS_ANALYZER_BASE_URL for LLM-driven page analysis; HTTP(S)_PROXY for network proxies. No unrelated credentials are requested.
Persistence & Privilege
always:false and the skill runs only when invoked. It stores browser login state under ~/.openclaw/xhs-browser (per SKILL.md/code), which is local to the user and expected for browser automation. The skill does not request elevated system-wide privileges or modify other skills' configurations.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install xiaohongshu-post-antaeus - 安装完成后，直接呼叫该 Skill 的名称或使用
/xiaohongshu-post-antaeus触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.7
修复：恢复稳定版本（模型驱动），修复图片上传、标题填写、正文写入、发布按钮点击等流程
v1.0.6
性能优化：减少模型调用从 6-10 次降至 2-3 次（硬编码固定流程+改进登录检测）
v1.0.5
性能优化：减少模型调用从 6-10 次降至 2-3 次（硬编码固定流程，只在关键节点调用模型确认）
v1.0.4
新增：使用 pyautogui 模拟 Esc 键关闭 macOS 系统文件选择对话框（需要辅助功能权限）
v1.0.3
修复：优化正文换行处理，使用<p>标签包裹段落，解决小红书编辑器过滤<br>导致内容连在一起的问题
v1.0.2
修复：优化正文换行处理，使用<p>标签包裹段落，解决小红书编辑器过滤<br>导致内容连在一起的问题
v1.0.1
修复：补充完整的环境变量声明（XHS_ANALYZER_BASE_URL, XHS_ANALYZER_MODEL, OPENAI_API_KEY, OPENAI_BASE_URL），解决安全扫描元数据不一致问题
v1.0.0
支持小红书笔记自动发布：浏览器自动化模式，自动上传图片、填写标题正文、点击发布
元数据
常见问题
Xiaohongshu Post (Browser Auto) 是什么？
Create and publish Xiaohongshu (小红书/RED) notes. Use when: user wants to publish a note to Xiaohongshu, schedule or batch post content, generate Xiaohongshu-s... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 259 次。
如何安装 Xiaohongshu Post (Browser Auto)？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install xiaohongshu-post-antaeus」即可一键安装，无需额外配置。
Xiaohongshu Post (Browser Auto) 是免费的吗？
是的，Xiaohongshu Post (Browser Auto) 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Xiaohongshu Post (Browser Auto) 支持哪些平台？
Xiaohongshu Post (Browser Auto) 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Xiaohongshu Post (Browser Auto)？
由 antaeus001（@antaeus001）开发并维护，当前版本 v1.0.7。
推荐 Skills