关于 Conqueror
Conqueror 为 Claude Code、Codex 及支持 SKILL.md 的 Agent 提供可复用 SEO 工作流。技能使用 Conqueror MCP 数据开展研究并提出页面修改建议;WordPress 改写会先以差异形式供审核。
主要功能
- 01研究并排列关键词机会
- 02聚类关键词并映射到页面
- 03分析竞争对手和市场格局
- 04将 WordPress 页面改写为待审差异
- 05寻找外链线索并起草联系内容
适用场景
- 01规划网站目标关键词
- 02将关键词组分配到现有或新页面
- 03研究竞争对手的排名与内容
- 04审核拟议的 WordPress 页面改写
MCP 相关
这些技能依赖文档所列远程端点上的 Conqueror MCP 服务来获取数据和工具。安装或运行技能前需先连接该服务。
配置
远程服务无需本地安装。选择你使用的客户端,复制配置或使用一键安装。
https://winwith.conquerorapp.com/mcpclaude mcp add --transport http com-conquerorapp https://winwith.conquerorapp.com/mcp{"mcpServers":{"com-conquerorapp":{"type":"http","url":"https://winwith.conquerorapp.com/mcp"}}}在 Cursor 中打开{"servers":{"com-conquerorapp":{"type":"http","url":"https://winwith.conquerorapp.com/mcp"}}}在 VS Code 中打开{"mcpServers":{"com-conquerorapp":{"type":"http","url":"https://winwith.conquerorapp.com/mcp"}}}{"mcpServers":{"com-conquerorapp":{"serverUrl":"https://winwith.conquerorapp.com/mcp"}}}命令在终端中运行;JSON 粘贴到对应客户端的 MCP 配置后重启。需要授权的服务会在首次连接时提示登录。
工具
README 原文English
来自项目 READMEREADME 暂无中文译文,以下为英文原文;中文介绍见上方。
Conqueror Agent Skills
SEO workflows for Claude Code, Codex, and other AI agents that support SKILL.md.
Each skill hands your agent one repeatable job — find keywords worth targeting, map them to pages, study a competitor, rewrite a page to win a term — using live data from Conqueror.
Connect Conqueror MCP first
These skills do nothing on their own. They tell your agent which workflow to run; the data comes from the Conqueror MCP server. Connect it before installing anything:
https://winwith.conquerorapp.com/mcp
In Claude Desktop or Claude Code, add that as a remote MCP server and sign in when prompted. You need a Conqueror account — start free.
Install
npx skills add Shipper-dot-now/conqueror-skills
Pick the skills and agents you want interactively, or take everything:
npx skills add Shipper-dot-now/conqueror-skills --skill '*'
Scope it to one agent with --agent claude-code or --agent codex.
To install by hand instead, copy the folders into your agent's skills directory:
git clone https://github.com/Shipper-dot-now/conqueror-skills.git
mkdir -p ~/.claude/skills && cp -R conqueror-skills/skills/* ~/.claude/skills/ # Claude Code
mkdir -p ~/.codex/skills && cp -R conqueror-skills/skills/* ~/.codex/skills/ # Codex
The skills
Start here if you're new:
| Skill | What it does |
|---|---|
/conqueror-seo-project-setup | Sets up a durable local workspace — working folder, site scope, goals, positioning, and Search Console intake — so your agent keeps context between sessions. |
/conqueror-seo-coach | Explains what each workflow does and recommends the next step. Use it when you're not sure what to run. |
Research:
| Skill | What it does |
|---|---|
/conqueror-keyword-research | Turns seed topics into a prioritized opportunity set, scored on volume, difficulty, intent, and SERP reality. |
/conqueror-keyword-clustering | Groups keywords into page-level clusters, maps each to an existing or new page, and flags cannibalization. |
/conqueror-competitive-landscape | Maps who wins a market, which content is working, and where the openings are. |
/conqueror-competitor-analysis | Deep dive on one competitor — ranking keywords, content themes, backlinks, weaknesses. |
Act on it:
| Skill | What it does |
|---|---|
/conquer-keyword | Rewrites your WordPress site's best-matching page for a target keyword and proposes it as a diff you review before anything goes live. Needs a WordPress connection. |
/conqueror-link-prospecting | Finds link prospects from SERPs and competitor backlinks, digs out public contact paths, and drafts the outreach. |
How they work
A skill is a markdown file. Your agent reads it when you run the matching slash command, then follows the workflow using the Conqueror MCP tools it describes. Nothing is compiled or executed — you can read every one of these files in a couple of minutes, and you should.
The skills never publish, apply, or spend on your behalf without asking. /conquer-keyword proposes a diff; applying it is your click in Conqueror.
To learn how skills work in your agent:
License
MIT — see LICENSE.
评论
一个开放协议,规定了 AI 应用如何以统一方式连接外部工具与数据源。客户端负责与模型对话,MCP 服务负责提供工具、资源和提示模板,两者通过标准消息通信,同一个服务可以被不同客户端复用。