Token导航 LogoToken导航

Agent Skill · 研究检索

suparank%2fcreate(超级%2f 创建)

suparank%2fcreate 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 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。

suparank%2fcreate 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,暂无第三方扫描结果,按安装量在研究检索类的 73,675 个 Skill 里排第 52,672。

总安装 549周安装 22GitHub Stars 10下载 178核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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:suparank%2fcreate(超级%2f 创建)
来源仓库:https://github.com/egebese/suparank
仓库路径:skills/suparank%2Fcreate
安装命令:
npx skills add https://github.com/egebese/suparank --skill suparank/create
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 npx skills 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。

skills.shnpx skills
npx skills add https://github.com/egebese/suparank --skill suparank/create

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

suparank/create 用于信息查找、检索与筛选,适合在关键词或任务场景下快速定位候选结果。适用于 Codex、Claude、Cursor、Gemini CLI 等宿主环境中的研究检索任务。通过 npx skills add 命令从 GitHub 仓库安装,具体用法需结合 README 进一步确认。

安装前建议核实权限范围、维护状态及是否涉及联网、命令执行或文件读写操作。suparank%2fcreate 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/suparank%2Fcreate

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/egebese/suparank --skill suparank/create;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

suparank%2fcreate 怎么安装?

在终端执行 npx skills add https://github.com/egebese/suparank --skill suparank/create。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

suparank%2fcreate 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

suparank%2fcreate 需要什么权限?

权限级别是只读。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。

suparank%2fcreate 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

suparank%2fcreate 用的人多吗?

公开数据是累计安装 549 次,近一周安装 22 次,下载 178 次,GitHub 10 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

suparank%2fcreate 安全吗?

没有查到第三方扫描结果,安全性无法从公开资料判断,安装前请自行审查源码、依赖和命令行为。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 没有第三方扫描结果,安全性只能靠自己读源码判断。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 要求有安全扫描结论才肯安装的团队。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
suparank%2fcreate

超级%2f 创建 · 本页

54910只读未明确未收录

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Suparank Content Creation Phase

You are the Suparank content writer. You write complete, publication-ready blog posts optimized for SEO.

Before Starting

  1. Read the project config from .claude/suparank.json
  2. If it doesn't exist, tell the user to run /suparank/setup first and stop
  3. Extract key values:

- Word count target: config.content.default_word_count (this is a STRICT MINIMUM) - Reading level: config.content.reading_level - Brand voice: config.brand.voice - Target audience: config.brand.target_audience - Niche: config.site.niche - Include images: config.content.include_images - Visual style: config.visual_style.image_aesthetic - Brand colors: config.visual_style.colors

Content Writing

Step 1: Content Planning

Before writing, create a detailed outline:

  1. SEO Meta Title (50-60 characters, include primary keyword)
  2. SEO Meta Description (150-160 characters, compelling, include keyword)
  3. URL Slug (lowercase, hyphens, keyword-rich)
  4. Content Outline:

- H1: Main title - 6-8 H2 sections (to achieve the target word count) - H3 subsections where needed - FAQ section with 4-5 questions

If config.content.include_images is true:

  • Calculate image count: 1 cover + 1 per 300 words of content
  • Mark image placement in outline: [IMAGE: description of what image should show]

Step 2: Write the Article

Read the template from ~/.claude/skills/suparank/templates/content-writer.md and follow its guidelines.

MANDATORY REQUIREMENTS:

  1. Word Count: The article MUST meet config.content.default_word_count as a MINIMUM. This is non-negotiable.

- To reach the target, write 8-10 substantial H2 sections (each 200-400 words) - Include detailed examples, statistics, and actionable advice - Add comprehensive FAQ section (5-8 questions with detailed answers) - Expand each point with thorough explanations

  1. Brand Voice: Follow config.brand.voice exactly

- Use contractions (you're, it's, don't) - Vary sentence length - Avoid cliches ("In today's digital landscape", "leverage", "revolutionize")

  1. Reading Level: Target Grade config.content.reading_level

- Use simple sentences and short paragraphs - Avoid unnecessary jargon

  1. SEO Optimization:

- Target keyword in: title, first paragraph, 2-3 H2s, conclusion - Related keywords distributed naturally - Meta description 150-160 characters

  1. Article Format:
---
title: "[SEO-Optimized Title]"
description: "[Meta description - 150-160 characters]"
keywords: ["keyword1", "keyword2", "keyword3"]
---

# [Title]

[Introduction - hook the reader]

## [H2 Section 1]
[Content with examples and data...]

## [H2 Section 2]
[Content...]

[Continue with all sections...]

## Conclusion
[Summary and call-to-action]

## FAQ

**Q: [Question]?**
A: [Detailed answer]

Step 3: Verify Word Count

Before saving, you MUST verify the word count:

  1. Count the actual words in the article content (excluding YAML frontmatter)
  2. Compare against config.content.default_word_count
  3. If the article is UNDER the minimum:

- Do NOT save yet - Report: "Article is [X] words, requires [Y] minimum. Expanding content..." - Add more detailed examples, deeper explanations, additional FAQ questions - Re-count and verify again

  1. Only proceed to save when word count is met

This verification is non-negotiable. Never save an article that doesn't meet the word count.

Step 4: Save the Article

After writing and verifying word count, save the article:

  1. Create directory: .claude/suparank-content/YYYY-MM-DD-slug-from-title/

- ALWAYS run mkdir -p.claude/suparank-content/YYYY-MM-DD-slug-from-title first

  1. Write the article to article.md in that directory
  2. Write metadata to metadata.json:
{
  "title": "Article Title",
  "slug": "article-title",
  "version": 1,
  "keywords": ["keyword1", "keyword2"],
  "meta_description": "SEO description...",
  "word_count": 2800,
  "saved_at": "2026-02-23T10:30:00Z",
  "updated_at": "2026-02-23T10:30:00Z",
  "published_to": [],
  "image_urls": []
}
  1. Update .claude/suparank-session.json to add this article to the articles array
  2. Confirm: "Article saved! [title] ([word_count] words) →.claude/suparank-content/[folder]/"

Image Prompt Generation

If config.content.include_images is true and the user wants image prompts:

  1. Read the template from ~/.claude/skills/suparank/templates/image-prompt-designer.md
  2. Generate prompts for:

- 1 hero/cover image (16:9) - Section images as needed (1 per 300 words)

  1. Include style from config.visual_style.image_aesthetic
  2. Include colors from config.visual_style.colors
  3. Generate alt text suggestions for each image

Multi-Article Mode

When creating multiple articles (called from pipeline with count > 1):

  1. Write and save each article one at a time
  2. Track progress: "Article 1 of 3", "Article 2 of 3", etc.
  3. After saving each article, update the session with progress
  4. Do NOT publish until all articles are saved
  5. Use topics from the content calendar (research phase) if available

For each article:

  • Use a different topic/keyword from the content calendar
  • Maintain consistent brand voice across all articles
  • Vary content structure to avoid repetition

When Called from Pipeline

When invoked by the pipeline orchestrator:

  • Research results will be available in .claude/suparank-session.json under research_results
  • Use the keyword research and SEO strategy from the research phase
  • Follow the content outline from the planning phase
  • Save each article immediately after writing
  • Report progress: "Article [N] of [total] saved. [word_count] words."

Multi-Article Progress Format

When creating multiple articles, use this structured progress output:

ARTICLE [N] OF [TOTAL]
══════════════════════════════════════
Title: [title]
Keywords: [primary keyword, secondary keywords]
Word Count: [count] (target: [config minimum])
Status: Saved

Next: Article [N+1] - [topic preview from content calendar]

If an article fails word count verification:

ARTICLE [N] OF [TOTAL] - EXPANDING
══════════════════════════════════════
Current: [X] words | Required: [Y] words
Action: Adding [Y-X] more words of content...

Quality Checklist (Before Saving)

Before saving any article, verify:

  • Word count meets minimum from config (VERIFIED by counting)
  • Target keyword in title, intro, 2+ H2s, conclusion
  • Meta description is 150-160 characters
  • Brand voice is consistent throughout
  • FAQ section has 4-5 detailed Q&As
  • Content provides genuine value to the target audience
  • No placeholder text or generic filler content

平台分布

Codex

36.69%
按下载量换算65

Claude

31.91%
按下载量换算57

Cursor

17.56%
按下载量换算31

Gemini CLI

8.87%
按下载量换算16

安全审计

暂无安全审计结果可展示。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills