← 返回 Skills 市场
mzcnyhhd

每日新闻总结

作者 MZCnyhhd · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
49
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install daily-news-summary
功能描述
每日从多个权威新闻来源(BBC RSS、南华早报、36氪、TechCrunch、The Verge、Microsoft/NVIDIA官方博客等)抓取最新资讯,自动生成四板块新闻日报:国际重要新闻、中国重要新闻、AI重要新闻、AI科技巨头动态。输出带原文链接的Markdown文件。当用户需要生成每日新闻摘要、科技资...
使用说明 (SKILL.md)

\r \r

Daily News Digest\r

\r 每日新闻日报生成器。从17个权威来源并行抓取新闻,整理为四板块结构化日报。\r \r

Workflow\r

\r

Step 1: Fetch Sources\r

\r 使用 web_fetch 工具并行抓取以下来源。RSS源优先,部分失败不影响整体流程。\r \r 国际新闻:\r

Step 2: Categorize Content\r

\r 从抓取结果中提取并分类到四板块:\r \r 板块一:国际重要新闻\r

  • 中东局势、地缘冲突、大国关系\r
  • 全球经济、金融市场、能源价格\r
  • 重大国际事件\r
  • 优先使用BBC RSS + 南华早报RSS\r \r 板块二:中国重要新闻\r
  • 宏观政策、外交动态\r
  • 财经数据、上市公司重要公告\r
  • 科技发射、社会热点\r
  • 南华早报RSS有大量大中华区深度报道\r \r 板块三:AI重要新闻\r
  • 大模型发布与迭代\r
  • AI产品与应用落地\r
  • 行业趋势与深度分析\r
  • 优先使用36氪RSS、TechCrunch RSS、The Verge RSS、量子位、机器之心\r \r 板块四:AI科技巨头动态\r
  • 按公司分类列出最新动态(OpenAI、Anthropic、Google、Microsoft、Meta、NVIDIA、DeepSeek、华为、字节、阿里等)\r
  • 每条动态附简要解读\r
  • 优先使用Microsoft Source RSS、NVIDIA Blog RSS、官方博客\r \r

Step 3: Generate Markdown Report\r

\r 文件格式要求:\r

  • 标题:每日新闻日报 | YYYY年MM月DD日\r
  • 标注生成时间和信息来源\r
  • 每条新闻末尾必须标注来源名称,并附上原文URL链接(格式:([来源名称](URL)))\r
  • 包含"今日最值得关注的Top 5事件"\r
  • 末尾包含"信息来源说明"表格\r
  • 语言:中文\r \r

Step 4: Save File\r

\r

  • 保存路径:工作区根目录\r
  • 文件名格式:新闻日报_YYYY-MM-DD.md\r \r

Optional: Automation\r

\r 用户可要求创建定时自动化任务(如"每天早上8点自动生成")。此时使用 WorkBuddy 的 automation_update 工具创建 recurring 任务,将上述 Workflow 作为自动化 prompt,schedule 设为 FREQ=DAILY;BYHOUR=8;BYMINUTE=0。\r \r

News Sources Reference\r

\r 详见 references/news_sources.md,包含17个来源的详细分类、URL、类型和可靠性评级。\r \r

Notes\r

\r

  • 如果某个网站抓取失败,继续用其他成功抓取的来源生成日报\r
  • BBC RSS、36氪RSS、TechCrunch RSS、The Verge RSS、南华早报RSS、NVIDIA Blog RSS、Microsoft Source RSS是目前最稳定可靠的来源\r
  • Reuters网页版和RSS经常失败,如果失败则依赖BBC+南华早报+新浪获取国际新闻\r
  • 联合早报RSS已停用,Google/Meta官方博客RSS暂不可用\r
  • 务必保留每条新闻的原文链接,方便用户点击阅读原文\r
  • 不要编造新闻,只基于实际抓取到的内容整理\r
  • Twitter/X API 可作为可选扩展(需用户自备 Bearer Token),本 Skill 不涉及\r
安全使用建议
This skill appears to do what it says (scrape public RSS/pages and produce a Markdown digest). Before installing: 1) Confirm what 'web_fetch' and 'WorkBuddy' tools are in your environment and what network/permission access they require — scheduling an automation will use WorkBuddy permissions. 2) Ensure you are comfortable with the agent writing files to the workspace root and with recurring scheduled tasks; prefer an isolated folder if concerned. 3) Verify scraping policy: the skill doesn't mention robots.txt, rate limits, or legal/copyright constraints — consider adding politeness (rate limiting) and user approval for any heavy scraping. 4) If you enable optional Twitter/X integration, only provide a token you trust and understand what it can access. 5) Monitor the first few runs to confirm sources and formatting, and revoke/disable the automation if outputs are unexpected.
功能分析
Type: OpenClaw Skill Name: daily-news-summary Version: 1.0.0 The skill is a standard news aggregator designed to fetch updates from 17 reputable sources (including BBC, TechCrunch, Microsoft, and NVIDIA) and generate a structured Markdown report. The workflow uses legitimate tools like `web_fetch` and `automation_update` for their intended purposes, and there is no evidence of data exfiltration, malicious execution, or harmful prompt injection in SKILL.md or the associated reference files.
能力标签
requires-oauth-tokenrequires-sensitive-credentials
能力评估
Purpose & Capability
The name/description promise (daily news digest from RSS and news sites) matches the runtime instructions: the SKILL.md lists public RSS and webpage URLs to fetch, categorization rules, and a Markdown output format. There are no unrelated environment variables, binaries, or install steps requested.
Instruction Scope
Instructions only ask the agent to use a web_fetch tool to retrieve listed RSS and webpages, extract headlines/content, categorize into four sections, include source links, and save a Markdown file at workspace root. They also describe creating a recurring automation via WorkBuddy's automation_update. The doc does not instruct reading other local files or credentials. Minor operational gaps: no mention of respecting robots.txt, rate limiting, or scraping constraints, and the workflow assumes web_fetch has permission to access those sites.
Install Mechanism
Instruction-only skill with no install spec and no code files — lowest risk from installation. Nothing is downloaded or written by an installer.
Credentials
The skill declares no required environment variables or credentials. It mentions optional use of Twitter/X API if the user supplies a Bearer Token, but explicitly states that the skill does not require it. One minor note: scheduling via WorkBuddy likely uses that tool's credentials or permissions (not declared here) — the skill itself doesn't request unrelated secrets.
Persistence & Privilege
always:false (normal). However, the skill explicitly instructs using WorkBuddy's automation_update to create recurring daily tasks; that creates persistent scheduled runs outside a one-off invocation. This is expected for an automation-oriented digest skill, but users should be aware creating recurring automations grants ongoing activity ability to the agent/tooling and may require separate approval in your environment.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install daily-news-summary
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /daily-news-summary 触发
  4. 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
v1.0.0
Initial release: 支持从17个权威来源自动生成四板块新闻日报(国际、中国、AI、科技巨头),带原文链接。
元数据
Slug daily-news-summary
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

每日新闻总结 是什么?

每日从多个权威新闻来源(BBC RSS、南华早报、36氪、TechCrunch、The Verge、Microsoft/NVIDIA官方博客等)抓取最新资讯,自动生成四板块新闻日报:国际重要新闻、中国重要新闻、AI重要新闻、AI科技巨头动态。输出带原文链接的Markdown文件。当用户需要生成每日新闻摘要、科技资... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 49 次。

如何安装 每日新闻总结?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install daily-news-summary」即可一键安装,无需额外配置。

每日新闻总结 是免费的吗?

是的,每日新闻总结 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

每日新闻总结 支持哪些平台?

每日新闻总结 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 每日新闻总结?

由 MZCnyhhd(@mzcnyhhd)开发并维护,当前版本 v1.0.0。

💬 留言讨论