Labsco
MCP SERVER

iReader MCP

by zlatanpham

Read the web back as text — a page, a YouTube transcript, a tweet thread, a PDF or a public Google Doc.

Page Fetching & Content Extraction
Summary
Five formats stop needing five different workarounds.

A link is a link — page, video, thread, PDF or doc — and each one has a tool that returns readable text, so research does not stall at whichever format the source happens to be in.

What it is

A reader for five kinds of internet content, each with its own tool. You pass a URL; what comes back is the content as text or Markdown rather than the page around it.

What you get
  • A webpage fetched through Jina reader and returned as Markdown
  • The transcript of a YouTube video, from its ID or URL
  • A tweet thread pulled back as a thread, not as a single post
  • The text of a PDF at a URL
  • A public Google Doc returned as Markdown
Requirements

Npx on your PATH — the published package runs without cloning anything. Every source is fetched live at request time, and the Google Doc tool only reaches documents that are publicly shared.

Setup effort

One command — npx -y @x-mcp/ireader@latest