aicreate360-reader-mcp

An MCP server that lets AI agents fetch any public webpage and convert it to clean Markdown.

Powered by aicreate360, it turns any URL into readable Markdown text — no manual copy-paste needed.

Works with:

• News sites, blogs, technical documentation
• GitHub pages, wikis
• Any publicly accessible webpage

Some websites are protected by anti-scraping measures or require authentication. These sites cannot be fetched, and the tool will return an appropriate error message.

中文版 README

Quick Start (npx)

No installation required — run directly with npx -y aicreate360-reader-mcp:

{
  "mcpServers": {
    "aicreate360": {
      "command": "npx",
      "args": ["-y", "aicreate360-reader-mcp"]
    }
  }
}

Add this to your MCP client config:

Client	Config path
Claude Code	~/.claude.json or .mcp.json
Claude Desktop	claude_desktop_config.json
Gemini CLI	~/.gemini/settings.json

Install via npm

npm i aicreate360-reader-mcp // in project
npm i -g aicreate360-reader-mcp // global

Then point your MCP client to the installed entry point:

{
  "mcpServers": {
    "aicreate360": {
      "command": "npx",
      "args": ["-y", "aicreate360-reader-mcp"]
    }
  }
}

Tool

This server exposes a single tool:

fetch_url

Fetch any public URL and return its content as clean Markdown.

Parameter	Type	Description
url	string (URL)	The public URL to fetch

Returns: Clean Markdown text.

Usage Examples

• Fetch the top 10 trending news from setn.com
• Convert this technical doc to Markdown: https://docs.github.com/en/rest
• Analyze this blog post: https://blog.example.com/post/1

Notes

• Each request fetches in real time — no caching.
• Depends on create360.ai service availability.
• No authentication required.
• Anti-scraping protected sites will return an error message.
• Requires Node.js >= 24.

License

MIT