Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问clear审计提醒

x-content-optimizerx 内容优化器

Agent Skill

用于辅助文档、README、Markdown、说明文和内容稿件的整理与改写。它适合让 Agent 提炼结构、补齐章节、统一术语、检查链接或把零散材料整理成可读文档。使用时应保留项目已有事实、命令和路径,不要把未确认的信息写成确定结论;涉及对外文案时,还需要控制语气,避免过度营销或夸大能力。

总安装

220

周安装

9

GitHub Stars

2

下载量

71
CodexClaudeCursorGemini CLI

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

3

许可证

MIT

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-content-optimizer(x 内容优化器)
来源仓库:https://github.com/newmanxbt/normans-skills
仓库路径:skills/x-content-optimizer
安装命令:
npx skills add https://github.com/newmanxbt/normans-skills --skill x-content-optimizer
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/newmanxbt/normans-skills --skill x-content-optimizer

简介

x-content-optimizer 用于辅助文档、README、Markdown 和内容稿件的整理与改写。

  • 适合让 Agent 提炼结构、补齐章节、统一术语或检查链接。
  • 通过 npx skills add 命令从指定 GitHub 仓库安装使用。
  • 需保留项目已有事实,避免写成确定结论;涉及对外文案时应控制语气,避免过度营销。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

X Optimizer

Audit content against X's recommendation algorithm (Phoenix/Grok-based) and provide optimization suggestions.

Workflow

Input Content → Algorithm Audit → Issue Report → Fact Check → Suggested Revision → User Approval → Final Output

Step 1: Receive Content

Accept content in any format:

  • Direct text paste
  • Markdown file
  • Multiple tweets (thread)

Identify content type:

  • Single Tweet: Under 280 chars, one post
  • X Article: Long-form native content
  • Thread: Multiple connected tweets

Step 2: Algorithm Audit

Reference: algorithm-rules.md

Audit Categories

A. Positive Signal Potential

SignalCheck
FavoriteEmotional resonance or valuable insight?
ReplyDiscussion hook or question?
RepostShare-worthy content?
DwellDepth to hold attention?
ClickCuriosity-inducing hook?

B. Negative Signal Risk

SignalCheck
Not InterestedRelevant to target audience?
Mute/BlockRespectful, not spammy?
ReportPolicy compliant?

C. Learned Penalties

FactorCheck
External LinksCore value IN content, not behind link?
ClickbaitPayoff matches promise?

D. Format & Structure

  • Hook strength (first line/50 chars)
  • Content completeness
  • Engagement driver (CTA/question)
  • Readability (line breaks, emphasis)

Step 3: Generate Audit Report

Format:

## Audit Report

### Score Summary
- Positive Signal Potential: [HIGH/MEDIUM/LOW]
- Negative Signal Risk: [HIGH/MEDIUM/LOW]
- Overall Algorithm Fit: [EXCELLENT/GOOD/NEEDS WORK/POOR]

### Issues Found
1. [Issue]: [Explanation]
   - Impact: [Which signal affected]
   - Fix: [Specific suggestion]

2. [Issue]: [Explanation]
   ...

### Strengths
- [What works well]

Step 4: Fact Check

CRITICAL: Before adding or suggesting any factual information (numbers, dates, statistics, claims about products/companies), you MUST:

  1. Identify factual claims in the content that need verification
  2. Search and verify using WebSearch tool
  3. Only include verified facts in the optimized version
  4. Remove or flag unverifiable claims - never fabricate data

Examples requiring verification:

  • Product launch dates ("X was released in...")
  • Statistics ("X% of users...")
  • Company announcements ("Company just launched...")
  • Performance metrics ("reduces time by X%")

If no verifiable source exists, either:

  • Remove the claim entirely
  • Replace with qualitative language ("significantly improves" instead of fake percentages)
  • Ask user if they have a source

Step 5: Generate Optimized Version

Provide complete rewritten version addressing all issues.

Format:

## Optimized Version

[Full rewritten content]

### Changes Made
1. [Change]: [Why it improves algorithm score]
2. [Change]: [Why it improves algorithm score]

Step 6: Offer One-Click Apply

If content is from a file:

Would you like me to apply these changes to the file?

If direct text:

Here's your optimized content ready to copy.

Content-Specific Guidelines

Single Tweet

  • Hook in first 50 chars (preview visibility)
  • Core value delivered completely
  • End with engagement driver
  • External links: provide context first

X Article

  • Title: curiosity + specific value
  • Opening: validate the click
  • Body: structured sections
  • Closing: summary + CTA
  • Advantage: keeps users on platform (dwell time)

Thread

  • Tweet 1: Standalone hook (must work alone)
  • Tweet 2-3: Engagement point (question)
  • Middle: Value delivery
  • Final: CTA (repost, follow)
  • Each tweet: Complete and independent

Quick Reference

High-Impact Improvements

  1. Add question/discussion hook → ↑ Reply
  2. Add quotable statement → ↑ Repost
  3. Strengthen opening hook → ↑ Click, Dwell
  4. Remove/contextualize external links → ↑ Dwell Time
  5. Add specific data/examples → ↑ Credibility, Share

Common Issues

IssueImpactFix
Weak hookLow click/dwellRewrite first line
Link-only valueUser leaves platformAdd value IN content
No engagement driverLow replyAdd question/CTA
Wall of textLow dwellAdd line breaks
Off-topic"Not interested" signalClarify audience fit

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

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

OpenCode

28.21%
按下载量换算20

windsurf

25.4%
按下载量换算18

Codex

18.29%
按下载量换算13

Claude Code

11.57%
按下载量换算8

Gemini CLI

8.32%
按下载量换算6

trae

3.75%
按下载量换算3

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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。

来源信息

继续浏览同类 Skills