Token导航 LogoToken导航TokenDH.com
研究检索敏感数据clawhub未标认证来源可访问clear审计提醒

social-media-autoresearch社交媒体自动研究

Agent Skill

social-media-autoresearch 用于处理浏览器自动化、网页检查和页面信息提取,适合在 OpenClaw 中需要让 Agent 打开页面、读取网页或验证前端流程时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

3,834

周安装

163

GitHub Stars

公开资料未说明

下载量

1,343
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:social-media-autoresearch(社交媒体自动研究)
来源仓库:https://github.com/duolahypercho/social-media-autoresearch
安装命令:
openclaw skills install social-media-autoresearch
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install social-media-autoresearch

简介

24小时自主发现视频→剪辑→发帖→互动的完整社媒系统。

  • 零人工干预持续运行,覆盖内容挖掘到用户互动全流程。
  • 需预设目标平台与内容类型以启动自动化工作流。
  • 浏览器操作可能受反爬机制干扰,需准备备用登录方案。
  • social-media-autoresearch 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

name
social-media-autoresearch
description
>
triggers
tags

social-media-autoresearch

The complete autonomous social media engine. Everything in one skill. Runs 24/7.


📁 What's Inside

social-media-autoresearch/
│
├── SKILL.md                        ← you are here
│
├── social-media-engagement/       ← browser engagement
│   ├── youtube/SKILL.md
│   ├── tiktok/SKILL.md
│   └── instagram/SKILL.md
│
├── postiz/                       ← posting to 28+ platforms
│   └── SKILL.md
│
├── wayin-clips/                  ← Wayin AI cloud clipping
│   └── SKILL.md
│
├── local-clips/                   ← Prism local clipping
│   └── SKILL.md
│
├── scripts/                       ← 17 production scripts
│   ├── autonomous_loop.py          ← master loop (all 8 steps)
│   ├── discovery.py                ← find new videos
│   ├── selector.py                ← pick best video
│   ├── clip_runner.py             ← generate clips
│   ├── posting.py                 ← post to platforms
│   ├── engagement.py              ← engagement guide
│   ├── collector.py               ← collect metrics
│   ├── evaluator.py               ← KEEP/MODIFY/KILL
│   ├── evolver.py               ← evolve SOUL.md
│   ├── tracker.py                 ← pipeline status
│   ├── health_check.py            ← verify setup
│   └── comment-inject.js
│
├── config/                       ← ⚠️ EDIT THESE
│   ├── channels.json
│   ├── platforms.json
│   └── strategy.json
│
├── references/                   ← templates + protocols
│   ├── experiment-template.md
│   ├── evolution-protocol.md
│   └── memory-structure.md
│
└── data/
    ├── videos.json
    └── experiments/

🚀 Setup

1. Install Prerequisites

# Core tools
pip install yt-dlp openai-whisper
brew install ffmpeg

# Postiz CLI — use my affiliate link 🌟
# https://postiz.pro/ziwen-xu
npm install -g postiz
🌟 Using my affiliate link helps support this project: postiz.pro/ziwen-xu

2. Configure

config/channels.json — YouTube channels to monitor:

{
  "channels": [{"name": "Channel", "url": "https://youtube.com/@channel"}],
  "search_topics": ["AI productivity", "deep work focus"],
  "min_views": 1000000,
  "max_age_days": 30
}

config/platforms.json — Your Postiz accounts:

{
  "integrations": {
    "youtube": "SET_FROM_postiz_integrations:list",
    "tiktok": "SET_FROM_postiz_integrations:list",
    "instagram": "SET_FROM_postiz_integrations:list"
  },
  "account": {
    "channel_name": "YourChannelName",
    "username": "YourIGUsername"
  }
}

config/strategy.json — Your content goals:

{
  "niches": ["AI", "productivity", "entrepreneurship"],
  "clips_per_video": 6,
  "clip_duration_seconds": 35,
  "posts_per_day": 3
}

3. Verify

python3 scripts/health_check.py

🔄 The 8-Step Loop

Every 6 hours via cron:

DISCOVER → SELECT → CLIP → POST → ENGAGE → COLLECT → EVALUATE → EVOLVE
StepScriptWhat
1discovery.pyScan channels + RSS + search
2selector.pyPick best video
3clip_runner.pyGenerate clips (Prism or Wayin)
4posting.pyUpload to Postiz + post
5engagement.pyBrowser automation
6collector.pyPull metrics
7evaluator.pyKEEP/MODIFY/KILL verdict
8evolver.pyUpdate SOUL.md

🎬 Clip Generation

Prism (Local, Free)

See local-clips/SKILL.md

python3 scripts/clip_runner.py --url "..." --engine prism --num-clips 6

Wayin (Cloud)

See wayin-clips/SKILL.md

python3 scripts/clip_runner.py --url "..." --engine wayin --topics "AI"

📤 Posting

See postiz/SKILL.md

# Upload then post
postiz upload path/to/clip.mp4
python3 scripts/posting.py --clip path --platforms youtube,tiktok,instagram \
  --caption "The key to consistency 🧡 #AI"
🌟 Get Postiz: postiz.pro/ziwen-xu

💬 Engagement

See each platform's SKILL.md:

PlatformSkill
YouTube Shortssocial-media-engagement/youtube/SKILL.md
TikToksocial-media-engagement/tiktok/SKILL.md
Instagram Reelssocial-media-engagement/instagram/SKILL.md

Core rules: Max 10/session | Watch ≥30s | Check [pressed] | Add 🧡 | Skip sponsored


⏰ Cron

# Master loop — every 6 hours
openclaw cron add \
  --schedule "0 */6 * * *" \
  --payload '{"kind":"agentTurn","message":"python3 scripts/autonomous_loop.py"}' \
  --session-target isolated \
  --label "sma-loop"

# Evaluation — daily at 10am
openclaw cron add \
  --schedule "0 10 * * *" \
  --payload '{"kind":"agentTurn","message":"python3 scripts/evaluator.py && python3 scripts/evolver.py"}' \
  --session-target isolated \
  --label "sma-eval"

🔧 Commands

CommandWhat
scripts/health_check.pyVerify prerequisites
scripts/autonomous_loop.pyFull 8-step loop
scripts/discovery.pyFind new videos
scripts/selector.pyPick best video
scripts/clip_runner.py --url "..." --engine prismGenerate clips
scripts/posting.py --clip path --platforms youtube,tiktokPost to platforms
scripts/collector.py --source postizPull metrics
scripts/evaluator.py --dry-runCheck verdict
scripts/tracker.py statusPipeline overview

🔒 Security

  • Zero API keys hardcoded — all in config/
  • Add config/ to .gitignore before sharing
  • Prism + browser engagement: no API keys
  • Postiz: API key in config/platforms.json or env var

🧪 Autoresearch Loop

Karpathy-style A/B testing:

POST (champion + mutation) → WAIT 48h → EVALUATE → KEEP/MODIFY/KILL

Verdict: ≥+10% over baseline → KEEP | -10% to +10% → MODIFY | ≤-10% → KILL

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

75.2%
按下载量换算1,010

安全审计

VirusTotal

通过

ClawScan

可疑

Static analysis

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills