RSS内容流引擎

v1.2.0

RSS 订阅 + AI 内容流引擎。监控多个 RSS 源,自动抓取最新文章, AI 分析提取核心观点,一键生成适配知乎/小红书/公众号的原创内容。 触发场景:用户说"帮我找今天的选题"、"RSS订阅"、"监控XX的更新"、 "把XX的文章改写成小红书"、"我需要一些内容灵感"、"自动生成今日内容"。 Keywor...

0· 154·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for freedompixels/rss-content-flow.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "RSS内容流引擎" (freedompixels/rss-content-flow) from ClawHub.
Skill page: https://clawhub.ai/freedompixels/rss-content-flow
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install rss-content-flow

ClawHub CLI

Package manager switcher

npx clawhub@latest install rss-content-flow
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The declared purpose (RSS monitoring, extracting core points, generating platform-specific drafts) matches the included scripts (manage_feeds.py and fetch_feed.py). Minor inconsistency: SKILL.md describes reading config from ~/.qclaw/skills/rss-content-flow/scripts/config.json, while the scripts read/write scripts/config.json relative to the package directory—this is a documentation/path mismatch but does not indicate malicious behavior.
Instruction Scope
Runtime instructions stay within the stated scope: read subscription config, fetch RSS XML, prepare structured JSON for AI analysis, and generate drafts. The SKILL.md references saving drafts via a 'feishu_doc' action and calling a 'social-media-poster' skill for publishing; those integrations are described but not implemented in the shipped scripts. No instructions ask the agent to read unrelated system files or environment variables.
Install Mechanism
No install spec or remote download is present. This is an instruction-and-script bundle only; nothing is fetched or installed during skill installation, so on-disk footprint is limited to the provided files.
Credentials
The skill declares no required environment variables, credentials, or special config paths. It performs network fetches of RSS feeds (expected). References to Feishu or social-media-poster imply external integrations, but the skill does not request or store credentials itself—those would be managed by the other skills if used.
Persistence & Privilege
The skill does not request always:true, does not modify other skills' configs, and only reads/writes a local config.json in its scripts directory. It does network I/O to public RSS feeds as expected; no elevated persistence or privileges are requested.
Assessment
What to consider before installing: - The code matches the RSS-fetching and feed-management description; it fetches RSS XML over HTTP(S) and writes a local scripts/config.json (not the ~/.qclaw path shown in SKILL.md). Expect network requests to the listed RSS URLs. - SKILL.md mentions saving drafts to Feishu and auto-publishing via a 'social-media-poster' skill. Those integrations are not implemented in the provided scripts, so verify how publishing and authorization will actually be handled before enabling automatic publish. - The skill does not request API keys or secrets itself; if you later enable auto-publish, confirm that the other skill (feishu/social-media-poster) only receives the minimum tokens/permissions required. - The path/documentation mismatch (SKILL.md vs scripts' CONFIG_FILE) is likely an editing oversight—confirm where your feed config will be stored and back it up if needed. - If you run the Python scripts locally, inspect/validate them (they use xml.etree.ElementTree to parse RSS and basic string cleaning) and consider running behind a network proxy if you require that for privacy/auditing. Overall this package appears coherent for its stated purpose; treat publishing integrations with normal caution and verify any external skill you authorize to post on your behalf.

Like a lobster shell, security has layers — review code before you run it.

latestvk972sdx4t3wq2hv3wj5nz4yhzh8508fs
154downloads
0stars
5versions
Updated 1w ago
v1.2.0
MIT-0

RSS 内容流

将 RSS 订阅转化为结构化内容素材,一站式生成多平台发布文案。

核心能力

  1. RSS 订阅管理 — 添加/删除/列出订阅源
  2. 智能抓取 — 自动获取最新 N 篇文章
  3. AI 分析 — 提取核心观点、关键数据、金句
  4. 内容生成 — 按平台风格改写(知乎/小红书/公众号)
  5. 草稿保存 — 保存到本地文件或直接输出

快速开始

用户说"帮我找今天的选题"时:

1. 读取 ~/.qclaw/skills/rss-content-flow/scripts/config.json(订阅列表)
2. 用 fetch_feed() 获取各源最新条目
3. AI 分析后呈现 3-5 个可用选题
4. 用户确认后 → 生成完整内容

订阅管理

添加订阅源

python3 scripts/manage_feeds.py --add <name> <url>
# 示例:python3 scripts/manage_feeds.py --add 36氪 https://36kr.com/feed

列出订阅

python3 scripts/manage_feeds.py --list

删除订阅

python3 scripts/manage_feeds.py --remove <name>

默认订阅源(预配置)

名称URL内容方向
36氪https://36kr.com/feed科技/商业
虎嗅https://www.huxiu.com/rss/0.xml商业/创业
少数派https://sspai.com/feed效率/工具
AI研习社https://ai.googleblog.com/feed/AI/技术
OpenAI Bloghttps://OpenAI.com/blog/rss.xmlAI产品

内容抓取

单源抓取

python3 scripts/fetch_feed.py --source 36氪 --limit 5

全源抓取

python3 scripts/fetch_feed.py --all --limit 3

输出示例:

📰 36氪 | 4篇新文章
  ① [AI创业] 标题:xxx | 浏览量预估:xxx
  ② [商业] 标题:xxx
  ...

📰 虎嗅 | 2篇新文章
  ① [汽车] 标题:xxx
  ...

抓取逻辑

  1. 解析 RSS XML,获取 title/link/description/pubDate
  2. 过滤:去除广告/软文/太旧的文章(>7天)
  3. 按时间倒序排列
  4. 返回结构化 JSON 供 AI 分析

AI 内容分析

收到文章列表后,AI 自动执行:

对每篇文章:
1. 读取 description / 摘要
2. 判断:这篇文章的核心价值是什么?
3. 提取:
   - 核心观点(1-2句)
   - 关键数据(数字/排名/增长)
   - 金句(适合引用的原话)
   - 相关话题标签(3-5个)
4. 评估:适合哪些平台?
   - 知乎:深度分析/观点争议类
   - 小红书:实操教程/工具推荐/个人经验类
   - 公众号:行业趋势/商业观察类

平台内容生成

知乎风格

  • 字数:800-1500字
  • 结构:背景→问题→分析→结论
  • 开头:痛点/争议性观点
  • 结尾:行动号召/评论区互动
  • 关键词:#AI #副业 #效率工具

小红书风格

  • 字数:300-600字
  • 结构:钩子→干货→互动
  • emoji 序号 + 话题标签(#AI副业 #效率神器)
  • 每段不超过3行
  • 结尾:评论区互动问题

公众号风格

  • 字数:1000-2000字
  • 结构:开篇引子→案例拆解→方法论→资源推荐
  • 标题公式:数字+情绪+悬念
  • 文末:往期回顾/互动话题

输出方式

保存为草稿(默认)

feishu_doc action=create title="内容草稿 | {日期}" → 获取 doc_token
feishu_doc action=write doc_token={token} content={内容}

预览后发布

生成内容 → 用户确认 → 调用 social-media-poster 发布

自动发布(需用户授权)

用户说"自动发布"时:
1. 将内容保存为本地Markdown文件
2. 调用 social-media-poster 发布
3. 返回发布结果

配置项

首次使用时向用户确认:

配置说明默认值
订阅源监控哪些 RSS36氪+虎嗅+少数派
抓取数量每次取几条每源3条
生成平台生成哪个平台内容用户指定
输出方式草稿/直接发布草稿
风格偏好正式/轻松/专业轻松

文件结构

rss-content-flow/
├── SKILL.md              # 本文件
├── README.md             # 用户文档
├── scripts/
│   ├── manage_feeds.py   # 订阅管理(增删查)
│   ├── fetch_feed.py     # RSS 抓取脚本
│   └── config.json       # 订阅配置(运行时生成)
├── references/
│   └── platform_style.md # 各平台风格指南
└── assets/               # 封面图模板等

与其他 Skill 的配合

  • social-media-poster:生成内容后直接发布到知乎/小红书
  • feishu-daily-report:每日内容草稿自动归入日报数据源
  • content-repurposer:同一篇文章改写到不同平台

注意事项

  • RSS 源需确保可访问,过期/失效的源自动跳过
  • 生成内容时必须注明原文来源(保留链接)
  • 版权合规:RSS 文章只能分析观点+改写,不得直接复制原文
  • 抓取间隔建议不低于 30 分钟,避免对源站造成压力

Comments

Loading comments...