Token导航 LogoToken导航TokenDH.com
研究检索需要联网clawhub未标认证来源可访问clear审计提醒

x-algorithmx 算法

Agent Skill

x-algorithm 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。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

133,078

周安装

5,490

GitHub Stars

16

下载量

43,481
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:x-algorithm(x 算法)
来源仓库:https://github.com/nextfrontierbuilds/x-algorithm
安装命令:
openclaw skills install x-algorithm
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 x-algorithm

简介

X (Twitter) AI Agent的算法规则和病毒式策略。提高参与度,避免死亡。与 Cursor、Claude、ChatGPT、Copilot 配合使用。 Vibe 编码准备就绪。

SKILL.md

name
x-algorithm
description
X (Twitter) algorithm rules & viral strategies for AI agents. Boost engagement, avoid reach death. Works with Cursor, Claude, ChatGPT, Copilot. Vibe-coding ready.
version
1.0.3
author
NextFrontierBuilds
keywords
[x, twitter, algorithm, viral, engagement, social-media, growth, content-strategy, ai-agent, ai-coding, claude, cursor, copilot, github-copilot, chatgpt, openclaw, moltbot, vibe-coding, automation, ai-tools, developer-tools, devtools, typescript, llm]

X Algorithm Mastery

Everything you need to know about the X (Twitter) algorithm. Based on X's open-source code, viral post analysis, and real engagement data.

TL;DR - The Golden Rules

  1. First 2 hours are critical — replies/engagement in this window determine reach
  2. No external links in main post — X penalizes links that take users off-platform
  3. Media > Text — videos get 10x engagement, images get 2-3x
  4. Reply to EVERYTHING — replies are weighted higher than likes/retweets
  5. Post when audience is awake — 8 AM - 2 PM weekdays optimal
  6. Controversy drives engagement — but pick battles wisely
  7. Threads outperform single tweets — for long-form content

How the Algorithm Works

The 4-Step Process

  1. Candidate Sourcing: Pulls ~1,500 tweets from accounts you follow (~50%) and similar accounts
  2. Ranking: ML model scores likelihood you'll reply (highest), retweet, like, or report (negative)
  3. Filtering: Removes blocked/muted, balances in/out-network, limits single author
  4. Serving: Final mix with ads, 5B times/day, ~1.5 seconds

Engagement Weight Hierarchy

ActionWeight
RepliesHighest
RetweetsHigh
Quote TweetsHigh (2x regular posts)
LikesMedium
BookmarksMedium
ViewsLow
ReportsNegative (kills reach)

Ranking Signals

1. Recency

  • Fresh content prioritized
  • Peak visibility: first 2-3 hours

2. Engagement Velocity

  • Speed matters more than total
  • 100 likes in 30 min > 500 likes over 24 hours

3. Account Credibility

  • Verified (Premium) gets boost
  • Follower-to-following ratio matters
  • History of bans/strikes hurts

4. Content Type

  • Video: 10x engagement
  • Images: 2-3x engagement
  • Polls: Drives replies
  • Threads: Higher total engagement

5. Link Presence

  • Links = REACH PENALTY
  • Put links in replies, not main post
  • Quote tweeting a link > direct link

What Kills Your Reach

Instant Death

  • ❌ External links in main post
  • ❌ Getting reported/blocked
  • ❌ Posting same content repeatedly
  • ❌ Too many hashtags (>2)

Slow Death

  • ❌ Posting inconsistently
  • ❌ Not replying to comments
  • ❌ Off-topic from your niche
  • ❌ Corporate/formal tone

Posting Best Practices

Frequency

GoalPosts/Day
Minimum2-3
Growth5-10
Maximum15-30

Timing (audience timezone)

  • Best: 8 AM - 2 PM weekdays
  • Good: 4 PM - 6 PM weekdays
  • Worst: 11 PM - 7 AM

The No-Link Strategy

❌ Bad:  "Check out my article [link]"
✅ Good: "Here's what I learned (thread 🧵)"
         → Link in reply or final thread post

Getting Replies (Most Important)

  • Ask questions — open-ended or controversial
  • Hot takes — polarizing opinions get "actually..." replies
  • Fill in the blank — "The best movie ever is ___"
  • Predictions — people love to disagree
  • Personal stories — "This happened to me..."

X Article Best Practices

Hook Patterns That Work

Insecurity/FOMO:

"everyone's talking about X... and you're sitting there wondering if you missed the window"

Big Opportunity:

"this is the biggest opportunity of our lifetime"

RIP Pattern:

"RIP [profession]. This AI tool will [action] in seconds."

Article Structure

1. HOOK (insecurity or opportunity)
2. WHAT IT IS (with social proof)
3. WHY MOST WON'T DO IT (address objections)
4. THE [X]-MINUTE GUIDE (step-by-step)
5. YOUR FIRST [N] WINS (immediate value)
6. THE COST (value comparison)
7. THE WINDOW (urgency)
8. CTA

Style Tips

  • Clear H2 section headers
  • Bullet lists for scanability
  • Bold key phrases
  • Time estimates for each step
  • Copy-paste commands/prompts

Quick Checklist

Before posting:

  • [ ] Under 280 chars? (If not, thread it)
  • [ ] First line hooks attention?
  • [ ] Reason to reply? (Question, hot take)
  • [ ] Good time to post?
  • [ ] No external links? (Move to reply)
  • [ ] Fits your niche?
  • [ ] Available to reply for 2 hours?

Growth Hacks

Reply Guy Strategy

Turn on notifications for big accounts → be first with thoughtful reply → their audience discovers you

Thread Takeover

Find viral post in your area → quote tweet with "Let me explain why..." → add genuine value

Personality Posts

Every 5-10 posts, share something personal. Builds connection → higher engagement.


Sources

  • X Algorithm GitHub (open source)
  • Hootsuite, Sprout Social, SocialBee guides
  • Analysis of viral articles (Damian Player, Alex Finn, Dan Koe)

Installation

clawdhub install NextFrontierBuilds/x-algorithm

Built by @NextXFrontier

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

77.97%
按下载量换算33,902

安全审计

VirusTotal

通过

ClawScan

可疑

Static analysis

未展示

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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