AI新闻多平台发布助手

每日抓取全球AI新闻，自动深度研究，生成高质量公众号长文并发布至微信和小红书

Security Scan

This is an instruction-only orchestrator that coordinates other skills to fetch AI news, research topics, generate long articles, create covers, and publish to WeChat and Xiaohongshu — its declared requirements and actions are coherent with its stated purpose.

Purpose & Capability

The name/description match what the SKILL.md and workflow.json describe: an orchestrator that calls ai-news-daily, agent-reach, 公众号助手, nano-banana-pro, wechat-mp-cn, and xiaohongshu-mcp to implement a multi-step publishing pipeline. It doesn't request unrelated credentials or binaries itself, which is proportionate for an orchestration skill.

Instruction Scope

The runtime instructions delegate actions (scraping Twitter/X, Reddit, YouTube, composing and publishing) to other named skills. That delegation is expected, but it means those downstream skills will perform network access and require credentials. The SKILL.md itself does not instruct reading local files, asking for secrets, or contacting hidden endpoints, which is appropriate; however, it encourages fully automated daily runs (cron) and publishing flows that may post content without human review if the agent is allowed to run autonomously.

Install Mechanism

No install spec and no code files — instruction-only — so nothing is downloaded or written to disk. This is the lowest-risk install profile and consistent with the skill being a workflow orchestrator.

Credentials

The skill declares no required environment variables, binaries, or config paths. That is proportionate because it only coordinates other skills; the credentials needed for posting/scraping should be requested by the respective platform-specific skills, not by this orchestrator.

Persistence & Privilege

always is false and disable-model-invocation is false (the default). The agent can invoke the skill autonomously, which is typical for skills, but combined with the skill's purpose (automatic publishing) the user should confirm whether autonomous execution and automatic posting are acceptable. The skill does recommend scheduling (cron), so users should ensure proper confirmation safeguards are in place.

Assessment

This skill is an orchestrator and appears coherent with its stated purpose, but before installing: (1) verify and inspect each downstream skill it calls (ai-news-daily, agent-reach, 公众号助手, nano-banana-pro, wechat-mp-cn, xiaohongshu-mcp) because those will request credentials and perform the actual scraping and publishing; (2) confirm posting behavior and require a manual review/approval step if you don't want automatic publication; (3) check that scraping/downstream skills comply with target platforms' terms of service and privacy rules; (4) be cautious with style imitation (e.g., "刘润式") to avoid impersonation or copyright issues; (5) test in a sandbox or staging account (not your primary WeChat/Xiaohongshu account) before enabling scheduled runs.

自动化的AI新闻内容工厂，每天从全球AI行业动态中抓取新闻，深度研究热点话题，生成3000-5000字专业公众号文章，并同时发布至微信公众平台与小红书。

技能协同图谱

[ai-news-daily]  →  抓取全球AI新闻
       ↓
[agent-reach]    →  深度研究热点话题（抓取Twitter/Reddit/YouTube评论）
       ↓
[公众号助手]      →  撰写专业深度文章（刘润风格，3000-5000字）
       ↓
[nano-banana-pro] →  生成封面配图
       ↓
┌──────────────┬──────────────┐
↓                              ↓
[wechat-mp-cn]          [xiaohongshu-mcp]
微信公众平台发布              小红书发布

工作流程

Phase 1: 新闻抓取

使用 ai-news-daily 技能获取当日全球AI行业最新动态，筛选出3-5条最有价值的新闻线索。

Phase 2: 深度研究

使用 agent-reach 技能对选定的新闻进行深度挖掘：

• 搜索Twitter/X上的专家评论和讨论
• 搜索Reddit上的社区反馈
• 搜索YouTube视频内容
• 综合多源信息，形成完整的故事脉络

Phase 3: 文章撰写

使用 公众号助手 生成专业长文：

• 字数：3000-5000字
• 风格：参考刘润公众号的专业商业分析风格
• 结构：背景→事件→分析→观点→结论
• 自动排版，不过度使用标题标签

Phase 4: 封面生成

使用 nano-banana-pro 生成封面图：

• 风格：专业、现代
• 尺寸：900×383px（公众号封面标准）
• 主题：与文章核心观点契合

Phase 5: 多平台发布

• 微信公众平台：使用 wechat-mp-cn 发布文章
• 小红书：使用 xiaohongshu-mcp 发布（精简版+封面图）

使用方式

【触发词】
"今日AI新闻" / "生成今日内容" / "发布公众号" / "更新内容"

【输入】
用户提供：文章主题方向（可选，默认AI行业热点）

【输出】
1. 今日AI新闻摘要
2. 完整公众号文章（Markdown格式）
3. 封面图片
4. 已发布的小红书帖子链接

适用人群

• 科技类自媒体作者
• AI行业从业者/研究员
• 内容创业者
• 需要每日持续输出内容的个人品牌

技术栈

• OpenClaw Agent
• ai-news-daily: 全球AI新闻抓取
• agent-reach: 多平台深度研究
• 公众号助手: 文章撰写与排版
• nano-banana-pro: 封面图生成
• wechat-mp-cn / xiaohongshu-mcp: 多平台发布