Token导航 LogoToken导航

Agent Skill · 研究检索

pre-publish-post-assistant(预发布后期助理)

pre-publish-post-assistant 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。

pre-publish-post-assistant 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,权限为需要联网;许可证为 MIT,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 41,921。

总安装 1,126周安装 46GitHub Stars 174下载 361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
MIT

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:pre-publish-post-assistant(预发布后期助理)
来源仓库:https://github.com/nicepkg/ai-workflow
仓库路径:skills/pre-publish-post-assistant
安装命令:
npx skills add https://github.com/nicepkg/ai-workflow --skill pre-publish-post-assistant
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/nicepkg/ai-workflow --skill pre-publish-post-assistant

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。

pre-publish-post-assistant 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。适用于预发布后期助理相关的关键词搜索和信息筛选场景。通过 npx skills add 命令从 GitHub 仓库安装并使用该技能。

安装前需确认权限范围和维护状态,注意可能触发联网或文件读写操作。建议结合原始 README 核验具体用法和功能边界。

来源仓库内路径:skills/pre-publish-post-assistant

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/nicepkg/ai-workflow --skill pre-publish-post-a…;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

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

pre-publish-post-assistant 怎么安装?

在终端执行 npx skills add https://github.com/nicepkg/ai-workflow --skill pre-publish-post-assistant。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

pre-publish-post-assistant 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。

pre-publish-post-assistant 需要什么权限?

权限级别是需要联网。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。

pre-publish-post-assistant 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

pre-publish-post-assistant 用的人多吗?

公开数据是累计安装 1,126 次,近一周安装 46 次,下载 361 次,GitHub 174 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。

pre-publish-post-assistant 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

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。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,商用和分发边界好判断。

先想清楚的地方

  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 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。

不太适合

  • 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
pre-publish-post-assistant

预发布后期助理 · 本页

1,126174需要联网MIT有提醒

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

Pre-publish Post Assistant Skill

Purpose

This skill helps prepare blog posts for publication by providing intelligent suggestions for:

  1. Categories - From existing site categories, with distribution awareness
  2. Tags - From existing tags, avoiding tag pollution
  3. SEO Metadata - Title, meta description, and focus keyphrase

All suggestions include rationale explaining the reasoning.

When to Use This Skill

  • User says "classify this post" or "suggest categories/tags"
  • User asks to "prepare this post for publishing"
  • User wants "SEO suggestions" for a draft
  • User provides a draft post and asks for taxonomy suggestions
  • User mentions "new blog post" and needs categorization help

Key Principles

Category Selection

  • Limit to 1-2 categories per post (primary + optional secondary)
  • Prefer categories with moderate post counts (avoid over/under-populated)
  • Match content theme, not just keywords
  • Consider category hierarchy if applicable

Tag Selection

  • Limit to 3-5 tags per post
  • Only use existing tags (no new tag creation unless explicitly requested)
  • Avoid tag pollution (tags with only 1-2 posts are low value)
  • Prefer tags that group related content meaningfully
  • Consider tag search potential

SEO Metadata

  • Title: 50-60 characters, include primary keyword, compelling
  • Meta Description: 150-160 characters, summarize value proposition, include CTA
  • Focus Keyphrase: 2-4 words, searchable, relevant to content

Input Formats

The skill accepts draft content in multiple formats:

# File path
"Classify this post: /path/to/draft.md"

# URL (for already-published posts needing optimization)
"Suggest tags for https://example.com/my-post/"

# Inline text
"Here's my draft: [content]... What categories fit?"

Data Sources

Categories and Tags

Can be retrieved from:

  1. WordPress GraphQL - Live data from WP
  2. Static dist folder - Parse from built site (/category/, /tag/ pages)
  3. Cached taxonomy file - Pre-generated taxonomy.json

Distribution Data

For balanced suggestions, the skill needs post counts per category/tag:

  • Categories: Aim for even distribution, flag if category would become oversized
  • Tags: Prefer tags with 5+ posts, warn about orphan tags

Output Format

## Suggested Categories

| Category | Post Count | Confidence | Rationale |
|----------|------------|------------|-----------|
| personal-development | 245 | High | Core theme matches self-improvement focus |
| productivity-effectiveness | 89 | Medium | Secondary theme around habits and routines |

**Recommendation**: Use "personal-development" as primary category.

---

## Suggested Tags

| Tag | Post Count | Confidence | Rationale |
|-----|------------|------------|-----------|
| habits | 45 | High | Central topic of the post |
| productivity | 67 | High | Directly discussed |
| morning-routine | 12 | Medium | Specific example in content |

**Recommendation**: Use all 3 tags. Avoid creating new tags.

---

## SEO Metadata

**Title** (58 chars):
> How to Build Morning Habits That Actually Stick | Your Blog

**Meta Description** (156 chars):
> Discover the science-backed approach to building morning habits that last. Learn the 3-step framework used by high performers. Start your transformation today.

**Focus Keyphrase**:
> morning habits

**Rationale**:
- "morning habits" has good search volume and matches user intent
- Title includes keyphrase naturally at the beginning
- Description creates urgency and promises specific value

Workflow

1. Analyze Content

  • Extract main themes and topics
  • Identify key concepts and terminology
  • Determine content type (how-to, opinion, review, etc.)

2. Load Taxonomy

  • Fetch existing categories with post counts
  • Fetch existing tags with post counts
  • Identify distribution patterns

3. Match & Score

  • Score each category/tag by relevance
  • Consider distribution balance
  • Flag potential issues (orphan tags, oversized categories)

4. Generate SEO

  • Craft title with primary keyword
  • Write compelling meta description
  • Suggest focus keyphrase

5. Present with Rationale

  • Show recommendations in table format
  • Explain reasoning for each suggestion
  • Highlight any concerns or alternatives

Configuration

{
  "taxonomy_source": "graphql|dist|file",
  "dist_path": "./dist",
  "taxonomy_file": "./taxonomy.json",
  "graphql_endpoint": "https://wp.example.com/graphql",
  "limits": {
    "max_categories": 2,
    "max_tags": 5,
    "min_tag_posts": 3
  },
  "seo": {
    "title_max_length": 60,
    "description_max_length": 160,
    "site_name": "Your Blog"
  }
}

Example Usage

Basic Classification

User: "Classify this post for me: /content/drafts/morning-routine-guide.md"

Claude: [Reads file, analyzes content, fetches taxonomy]
        [Presents category/tag suggestions with rationale]
        [Generates SEO metadata]

Quick Tag Check

User: "What tags should I use for a post about Bitcoin ETFs and institutional adoption?"

Claude: [Analyzes topic, checks existing tags]
        "Based on your existing tags, I recommend:
        - bitcoin (89 posts) - primary topic
        - cryptocurrency (45 posts) - broader category
        - investing (23 posts) - relevant angle

        Avoid creating new tags like 'etf' or 'institutional' unless you plan
        to write more content on these specific topics."

SEO Focus

User: "Generate SEO metadata for my post about productivity apps for remote workers"

Claude: [Analyzes topic and search intent]

        Title: "Best Productivity Apps for Remote Workers in 2025 | Your Blog"

        Description: "Discover the top productivity apps that remote workers
        swear by. From task management to focus tools, find the perfect
        stack for your home office."

        Focus Keyphrase: "productivity apps remote workers"

Best Practices Enforced

  1. No tag pollution - Won't suggest creating new tags unless justified
  2. Balanced distribution - Warns if category is becoming oversized
  3. SEO compliance - Enforces character limits and keyword placement
  4. Existing taxonomy - Always checks against actual site data
  5. Transparent reasoning - Every suggestion includes rationale

平台分布

Claude Code

31.54%
按下载量换算114

OpenCode

23.54%
按下载量换算85

Cursor

15.88%
按下载量换算57

trae

11.27%
按下载量换算41

Gemini CLI

8.6%
按下载量换算31

Codex

3.34%
按下载量换算12

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。

继续浏览同类 Skills