X Engagement
X/Twitter 运营自动化。完整 onboarding → Persona 学习 → 人类行为模拟 → 记忆系统 → 定时任务 → For You 关注 → Following 互动 → 自我进化系统
MIT-0 · Free to use, modify, and redistribute. No attribution required.
⭐ 1 · 239 · 1 current installs · 1 all-time installs
byJason@jasonCodeSpace
MIT-0
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Name/description (X engagement automation) match the delivered artifacts: browser DOM scripts, persona/memory system, comment-generation, and cron job docs. No unrelated credentials or binaries are requested; the skill intentionally uses browser automation (profile=openclaw) instead of API tokens, which is coherent with its stated approach.
Instruction Scope
Runtime instructions and provided scripts instruct the agent to read and write local memory files (memory/daily/hotspots/*), generate and post comments via evaluated browser JS, and check/modify cron tasks. These are within the skill's purpose but broaden its scope to system file I/O and scheduling. The skill also directs DOM actions designed to evade detection (human behavior simulation), which is deceptive by design and increases risk to user accounts.
Install Mechanism
There is no network install/download; this is instruction‑plus‑scripts (no external archive URLs). That lowers remote code-fetch risk. However, the repository includes executable shell scripts that will be placed on disk and can be run by setup scripts—so local execution risk remains and should be inspected.
Credentials
The skill requests no environment variables, no API keys, and no external service credentials. It relies on a logged-in browser session (user account) rather than asking for tokens, which is consistent with the design. There are no obvious unrelated secrets requested.
Persistence & Privilege
The skill provides scripts (scripts/setup-cron.sh, scripts/cleanup-memory.sh, scripts/check-cron.sh) and explicit instructions to add cron jobs (via openclaw cron or crontab). That means it will create persistent scheduled tasks that run autonomously and modify the user's crontab and files under ~/memory or relative memory/ paths. While not 'always:true', this persistent system-level presence and self-scheduling increases blast radius and should be consciously approved by the user.
What to consider before installing
This skill is internally consistent as a Twitter/X automation bot, but it performs actions with real risk: it will read/write local memory files, execute scripts, and create cron jobs that autonomously drive a browser to like, follow, and post comments (with behavior intended to evade detection). Before installing: 1) Review scripts setup-cron.sh and cleanup-memory.sh line-by-line and confirm any crontab edits; 2) back up your crontab and run setup scripts manually (or not at all) until you audit them; 3) run the skill in an isolated/test account or sandboxed machine (not your main account) because posting automation can lead to account suspension or policy violations; 4) confirm you have an explicit, logged-in browser profile for the target account (the skill uses browser automation, not API tokens); 5) consider disabling automatic cron setup and run tasks manually until you trust the behavior; 6) if you are not comfortable with scheduled autonomous posting or with scripts modifying crontab/files under your home directory, do not install. If you want a safer review, paste the exact contents of scripts/setup-cron.sh and scripts/cleanup-memory.sh here and I can point to any exact commands that will modify system state.Like a lobster shell, security has layers — review code before you run it.
Current versionv4.1.2
Download ziplatest
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
SKILL.md
X 运营自动化 Skill v4.0
快速开始
触发条件:
- "刷推 [时间]"
- "运营推特 [时间]"
- "去X上互动 [时间]"
首次运行:
自动进入 Onboarding 流程(详见 docs/onboarding.md)
后续运行: 读取配置 → 直接刷推
文档结构
x-engagement/
├── SKILL.md # 主入口(本文件)
├── docs/
│ ├── onboarding.md # Onboarding 流程
│ ├── browser-operations.md # 浏览器操作模块(基于 Chirp)
│ ├── comment-rules.md # 评论规则(重要!防止错误)
│ ├── human-behavior.md # 人类行为模拟规范
│ ├── memory-system.md # 记忆系统设计
│ ├── cron-jobs.md # 定时任务
│ ├── comment-generation.md # 评论生成逻辑
│ └── natural-language-parser.md # 自然语言时间解析
├── playbooks/
│ ├── comment-strategies.md # 评论策略(有效/无效)
│ └── changelog.md # 策略变更记录
├── data/
│ └── engagement/
│ └── YYYY-MM-DD.json # 每日评论数据
├── templates/
│ ├── persona.md # Persona 模板
│ ├── config.json # 配置模板
│ └── daily-log.md # 每日日志模板
└── scripts/
├── setup-cron.sh # 设置定时任务
├── check-onboarding.sh # 检查状态
└── daily-review.sh # 每日复盘脚本
核心功能
1. Onboarding(首次运行)
5个阶段:
- 浏览器连接 + 登录检查
- 选择 Persona(自己或其他账号)
- 学习 Persona(抓取100条 → 生成描述)
- 刷推习惯配置
- 保存配置
详见: docs/onboarding.md
2. 人类行为模拟
核心原则: 不追求完美,追求"足够像真人"
包含:
- 随机时间生成器(正态分布)
- 人类滚动模式(小/中/大滚动)
- 鼠标轨迹模拟
- 频率限制
- 评论间隔(3-6分钟)
详见: docs/human-behavior.md
3. 记忆系统
三层记忆:
memory/daily/hotspots/
├── .onboarding_complete # Onboarding 标记
├── .config.json # 用户配置
├── personas/
│ └── [handle].md # Persona 描述
├── events/ # 重大事件(永久)
├── tables/ # 每日热点(7天)
└── history/
├── comments/ # 评论历史(避免自相矛盾)
└── daily/ # 每日日志
关键功能:
- 记录每次评论内容
- 记录用户说过的话(如"昨天出去吃饭了")
- 评论前检查历史,避免矛盾
详见: docs/memory-system.md
4. 定时任务
每日热点总结:
- 时间:每天早上10点
- 内容:抓取 Top 10 → 更新热点表格 → 推送给用户
刷推提醒(用户自定义):
- 支持自然语言设置
- 固定时间:"早上9点、下午3点、晚上9点"
- 随机时间:"每天3次,随机时间"
- 工作日/周末:"工作日晚上8点,周末随机3次"
设置方法:
./scripts/setup-cron.sh
详见: docs/cron-jobs.md
5. 刷推流程
⚠️ 重要规则(必须遵守):
- 只在 Following 的 Recent 页面评论(不是 Popular)
- 评论前检查历史(避免重复评论同一博主)
- 记录所有评论(保存到历史文件)
详见: docs/comment-rules.md(必读!)
For You 页面:
- 浏览(真人滚动模式)
- 关注(根据配置条件)
Following 页面:
- 确保是 Recent(不是 Popular)
- 点赞(有价值的推文)
- 评论(2小时内,使用 persona 风格)
- 记录评论到历史(避免重复)
详见: docs/comment-generation.md
6. 浏览器操作(基于 Chirp)
使用 profile=openclaw:
- 独立浏览器进程,100% 稳定
- 不依赖 Chrome 扩展(Browser Relay 不稳定)
- 支持 DOM 操作 + 人类行为模拟
核心操作:
- 点赞(带随机延迟)
- 评论(使用 persona 风格)
- 滚动(模拟人类)
- 关注(根据条件)
Token 优化:
- 避免频繁 snapshot
- 使用 compact snapshot
- 直接操作 DOM
详见: docs/browser-operations.md
使用示例
首次使用
用户: 刷推
Bot: 开始 Onboarding...
1. 检查浏览器...
2. 请选择 persona...
3. 学习中...
4. 配置刷推习惯...
5. 完成!开始刷推...
后续使用
用户: 刷推半小时
Bot: 读取配置...
For You: 浏览 18-30 屏,关注 6 人
Following: 点赞 18 条,评论 6 条
开始...
关键特性
| 特性 | 说明 |
|---|---|
| 完整 Onboarding | 5阶段引导,学习 persona |
| 人类行为模拟 | 随机延迟、滚动模式、频率限制 |
| 记忆系统 | 评论历史、用户信息、热点表格 |
| 定时任务 | 每日热点总结 |
| 避免矛盾 | 评论前检查历史记录 |
| 结构化设计 | 多文件组织,易于维护 |
自我进化系统
核心理念
没有记忆的AI,只是一个聪明的工具。 有记忆且能进化的AI,才是会成长的伙伴。
进化闭环
采集数据 → 分析对比 → 得出结论 → 更新规则 → 下次执行
三大机制
1. Playbook 系统
playbooks/comment-strategies.md- 记录有效/无效策略playbooks/changelog.md- 记录策略变更- Agent 可以更新自己的规则
2. 数据采集
data/engagement/YYYY-MM-DD.json- 每日评论数据- 记录:时间、作者、内容、结果
- 用于后续分析和优化
3. 每日复盘(22:00)
- 统计今日数据
- 分析有效策略
- 更新 Playbook
- 生成明日建议
- 推送报告给用户
文件结构
x-engagement/
├── playbooks/
│ ├── comment-strategies.md # 评论策略(有效/无效)
│ └── changelog.md # 策略变更记录
├── data/
│ └── engagement/
│ └── YYYY-MM-DD.json # 每日评论数据
└── scripts/
└── daily-review.sh # 每日复盘脚本
使用示例
Agent 学习过程:
- 发现「妙啊」评论效果好
- 在 Playbook 中记录:「妙啊」适用于技术分享,数据支撑:2026-03-02
- 下次刷推时读取这条规则
- 考虑在类似推文上使用相同策略
进化效果:
- Agent 越用越聪明
- 自动学习什么评论有效
- 持续优化策略
- 避免重复错误
必读文档
按顺序阅读:
docs/onboarding.md- 了解首次运行流程docs/human-behavior.md- 了解人类行为模拟docs/memory-system.md- 了解记忆系统docs/comment-generation.md- 了解评论生成
版本: 4.0.0 更新: 2026-03-02 改进: 结构化设计 + 记忆系统 + 定时任务 + 人类行为规范
Files
19 totalSelect a file
Select a file to preview.
Comments
Loading comments…
