Token导航 LogoToken导航TokenDH.com
研究检索external-servicegithub未标认证来源可访问许可证需确认审计提醒

wordpress-mcpWordPress MCP 搜索

Agent Skill

wordpress-mcp 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。

总安装

559

周安装

24

GitHub Stars

2

下载量

196
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

2

许可证

unknown

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:wordpress-mcp(WordPress MCP 搜索)
来源仓库:https://github.com/5unnykum4r/wordpress-mcp
仓库路径:skills/wordpress-mcp
安装命令:
npx skills add https://github.com/5unnykum4r/wordpress-mcp --skill wordpress-mcp
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/5unnykum4r/wordpress-mcp --skill wordpress-mcp

简介

用于通过 MCP 协议管理 WordPress 站点的工具集。

  • 适合批量操作文章、媒体或站点配置的远程维护任务。
  • 依赖本地 PHP 适配器与 Application Password 认证机制。
  • 部署前需确认服务器安全策略与 API 访问权限边界。
  • wordpress-mcp 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

WordPress MCP Site Management

Manage WordPress sites through a single Python MCP proxy server (wordpress in ~/.claude.json). All tools require a site parameter identifying the target site. Sites are configured in sites.json.

Setup

This skill requires the WordPress MCP server. If the wordpress MCP tools are not yet available, set up the full project from https://github.com/5unnykum4r/wordpress-mcp:

  1. Clone the repo and install the PHP adapter on your WordPress site (wordpress/ directory)
  2. Create a WordPress Application Password and add your site to sites.json
  3. Add the MCP server to ~/.claude.json:
{
  "mcpServers": {
    "wordpress": {
      "type": "stdio",
      "command": "uv",
      "args": ["run", "--with", "fastmcp", "--with", "httpx", "python", "/path/to/wordpress-mcp/server.py"]
    }
  }
}

See SETUP.md in the repo for the full step-by-step guide.

How It Works

A Python FastMCP server proxies tool calls to any WordPress site:

  1. Every tool takes site as the first parameter (e.g. site="myblog")
  2. The proxy looks up credentials in sites.json
  3. Sends JSON-RPC requests to the WordPress MCP adapter endpoint
  4. Authentication uses WordPress Application Passwords via HTTP Basic Auth

Use list_sites to see all configured sites.

Available Tools

See references/posts.md, references/categories-tags.md, references/media.md, references/comments.md, references/redirections.md, references/admin.md, and references/tablepress.md for full parameter details.

Quick Reference

All tools require site as the first parameter.

ToolActionKey Params
list_postsList posts/pagesstatus, category, tag, search, post_type, page
read_postRead full post + SEOpost_id
create_postCreate post/pagetitle, content, slug, status, author, categories, tags, rank_math
update_postUpdate post/pagepost_id + any fields to change (incl. author)
delete_postTrash/deletepost_id, permanent
bulk_update_statusBulk status changepost_ids[], status
list_revisionsRevision historypost_id
restore_revisionRestore revisionrevision_id
search_replaceFind/replace in postssearch, replace, dry_run
list_categoriesList categories + SEOhide_empty, parent, search
create_categoryCreate categoryname, slug, description, parent_id, rank_math
update_categoryEdit categorycategory_id + fields, rank_math
delete_categoryDelete categorycategory_id
list_tagsList tagshide_empty, search
create_tagCreate tagname, slug, description
update_tagEdit tagtag_id + fields
delete_tagDelete tagtag_id
upload_imageUpload from URLimage_url, title, alt_text, filename
list_mediaBrowse media librarymime_type, search, page
update_mediaEdit media metadataattachment_id, title, alt_text, caption
delete_mediaDelete media itemattachment_id
list_commentsList commentspost_id, status, search
update_commentApprove/spam/replycomment_id, status, reply
delete_commentDelete commentcomment_id
list_redirectionsList Rank Math redirectssearch
create_redirectionCreate redirectfrom_url, to, type (301/302)
update_redirectionEdit redirectredirection_id + fields
delete_redirectionDelete redirectredirection_id
list_pluginsList installed plugins
toggle_pluginActivate/deactivateplugin, action
list_usersList usersrole, search
manage_optionsRead/write WP settingsread[], write{}
clear_cacheFlush all caches
get_infoSite info + stats
list_block_typesList Gutenberg blocksnamespace, custom_only
list_patternsList synced patternssearch
read_patternRead pattern contentpattern_id
create_patternCreate synced patterntitle, content
update_patternUpdate patternpattern_id + fields
delete_patternDelete patternpattern_id
list_tablepress_tablesList TablePress tables
read_tablepress_tableRead table data + optionstable_id
create_tablepress_tableCreate new tablename, data[][], options
update_tablepress_tableUpdate tabletable_id + any fields
delete_tablepress_tableDelete tabletable_id
list_sitesShow configured sites

Rank Math SEO

Posts and categories support a rank_math object on create/update.

Post fields:

title, description, focus_keyword, canonical_url, robots[],
primary_category, pillar_content, breadcrumb_title, schema_type,
og_title, og_description, og_image, twitter_title, twitter_description

Category fields:

title, description, focus_keyword, canonical_url, robots[],
breadcrumb_title, og_title, og_description, og_image,
twitter_title, twitter_description

read_post and list_categories return all Rank Math fields.

Gutenberg Blocks

Post content is stored as Gutenberg block markup. read_post returns the raw block HTML including <!-- wp:block-name --> delimiters. create_post and update_post accept full block markup in the content field — block comments are preserved as-is.

Use list_block_types to discover all registered blocks (core + custom). Pass custom_only=True to see only theme/plugin blocks, or namespace="your-theme" to filter by namespace. Each block entry includes its attribute schema so you can construct valid block markup.

Block markup format:

<!-- wp:namespace/block-name {"attribute":"value"} -->
<div class="wp-block-namespace-block-name">Inner HTML</div>
<!-- /wp:namespace/block-name -->

When updating an existing post's content, always read_post first to get the full block markup, modify the specific blocks, then pass the entire content back.

Common Workflows

Create an optimized post

  1. upload_image(site="myblog", image_url="...") — upload featured image
  2. create_post(site="myblog", title="...", content="...", categories=["Tech"], rank_math={...}) — create with SEO
  3. read_post(site="myblog", post_id=123) — verify
  4. update_post(site="myblog", post_id=123, status="publish") — publish

Bulk SEO audit

  1. list_posts(site="myblog", number=50) — get all published posts
  2. read_post(site="myblog", post_id=X) — check rank_math fields
  3. update_post(site="myblog", post_id=X, rank_math={...}) — fix missing SEO

Manage redirects after slug change

  1. read_post(site="myblog", post_id=123) — get current slug
  2. update_post(site="myblog", post_id=123, slug="new-slug") — change slug
  3. create_redirection(site="myblog", from_url="/old-slug", to="/new-slug", type=301) — redirect

Search and replace

  1. search_replace(site="myblog", search="old text", replace="new text", dry_run=True) — preview
  2. search_replace(site="myblog", search="old text", replace="new text", dry_run=False) — execute

Create a reusable synced pattern

  1. create_pattern(site="myblog", title="CTA Banner", content="<!-- wp:group -->...") — returns ref block markup
  2. Use the returned ref_block (e.g. <!-- wp:block {"ref":456} /-->) in any post content
  3. Updating the pattern auto-updates every post that references it

Create a post with custom Gutenberg blocks

  1. list_block_types(site="myblog", custom_only=True) — discover custom blocks and their attributes
  2. Build block markup: <!-- wp:namespace/block-name {"attr":"value"} -->\n<div>...</div>\n<!-- /wp:namespace/block-name -->
  3. create_post(site="myblog", title="...", content="...block markup...", rank_math={...}) — create post
  4. read_post(site="myblog", post_id=123) — verify block content preserved

Edit blocks in an existing post

  1. read_post(site="myblog", post_id=123) — get current block markup from content
  2. Modify the specific <!-- wp:... --> blocks in the content string
  3. update_post(site="myblog", post_id=123, content="...updated full content...") — save back

Create and embed a TablePress table

  1. create_tablepress_table(site="myblog", name="Pricing", data=[["Plan","Price"],["Basic","$9"],["Pro","$29"]]) — returns id + shortcode
  2. update_post(site="myblog", post_id=123, content="... [table id=5 /]...") — embed in post
  3. read_tablepress_table(site="myblog", table_id="5") — verify

Managing Sites

Sites are configured in sites.json:

{
  "myblog": {
    "url": "https://example.com/wp-json/mcp/mcp-adapter-default-server",
    "username": "AdminUsername",
    "password": "xxxx xxxx xxxx xxxx xxxx xxxx"
  }
}

To add a new site: install the MCP adapter on the WordPress server (see SETUP.md), create an Application Password, and add an entry to sites.json. No restart needed.

Site-Specific Extensions

For sites with custom workflows (e.g. ACF custom fields, specific content templates, SEO recovery plans), create additional skill files alongside this one and reference them from your project's CLAUDE.md or .claude/skills/ directory.

Example structure:

~/.claude/skills/
├── wordpress-mcp/         ← This skill (generic WordPress tools)
│   ├── SKILL.md
│   └── references/
└── my-site-workflow/      ← Your custom skill (site-specific context)
    └── SKILL.md

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

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

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

Codex

38.31%
按下载量换算75

Claude

26.61%
按下载量换算52

Cursor

17.03%
按下载量换算33

Gemini CLI

8.92%
按下载量换算17

安全审计

Gen Agent Trust Hub

通过

Socket

可疑

Snyk

可疑

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

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