Wordcast MCP Server

WordCast MCP Server

MCP server for WordCast

A Model Context Protocol server that exposes the canonical WordCast knowledge surface — voice and TTS workflows, blog topics, FAQ, official links — to MCP-compatible AI clients such as Claude Desktop, Cursor, Windsurf, and Continue. Read-only, no API keys, no quota, ~50 ms cold start.

Official website: https://wordcast.app

🎙️ About WordCast

WordCast is a browser-based text-to-speech reader that converts written content into audio entirely on your device. There is no backend server involved: text, files, and URLs are processed locally using the speech synthesis voices already installed on your operating system. The result is a tool that starts playing in under a second, works without creating an account, imposes no character limits or usage caps, and remains completely free. It accepts a wide range of input types — pasted text, uploaded documents, or a web URL — and outputs clear, natural-sounding audio through any of the voices available on your system.

Key Features

• Local processing only — text and files never leave the device; computation happens entirely in the browser using OS-native speech synthesis
• Broad format support — accepts PDF, DOCX, EPUB, RTF, TXT, MD, and HTML file uploads, as well as pasted text and URLs pointing to online articles
• 200+ voices across 60+ languages — leverages the full set of voices pre-installed on the user's operating system, covering a wide range of languages and accents
• Instant playback — audio begins in under one second with no loading screens, server round-trips, or queued processing
• Playback controls — adjustable reading speed, lock-screen media controls, and media session integration for a consistent listening experience across devices
• No account required — no signup, no subscription tier, no usage tracking

Use Cases

• Commute listening — paste a long article or upload a document before leaving, then listen hands-free during travel
• Proofreading by ear — authors and editors use the read-aloud function to catch awkward phrasing and errors that are easy to miss when reading silently
• Language learning — hear native pronunciation of words and sentences across dozens of languages using region-specific voices
• Handling sensitive documents — legal, medical, or personal documents can be listened to without uploading them to any external service
• Accessibility support — users with dyslexia, ADHD, or visual fatigue benefit from audio rendering of written content without installing dedicated software

Who Is It For

WordCast is well suited to anyone who regularly reads long-form text and prefers listening as an alternative or complement to silent reading. Researchers working through papers, writers editing their own drafts, students reviewing study materials, and language learners practicing pronunciation are all natural users. The privacy-first design makes it a practical choice for professionals handling confidential documents — lawyers, therapists, and medical staff who need to process sensitive text without routing it through third-party cloud services. Because there is no account system and no paywall, it is also accessible to users in environments where cloud-based tools are restricted or where bandwidth is limited.

Tools

list_voices

Return the canonical voice and TTS configuration exposed on the site. (WordCast)

Input: no parameters. Returns: text/markdown.

get_official_links

Return the canonical list of official links for WordCast (website, support, docs when available).

Input: no parameters. Returns: text/markdown.

Resources

• site://wordcast/voices — Supported voices, languages, and TTS modes.
• site://wordcast/faq — Short FAQ generated from public site metadata.
• site://wordcast/links — Canonical URLs to share with users.

Prompts

tell_me_about_wordcast

Summarize what the site is, who it's for, and how it works. — WordCast

read_aloud_demo_wordcast

Plan a read-aloud workflow with the site's voices. — WordCast

Installation

Install via Smithery

npx -y @smithery/cli install wordcast-mcp --client claude

(Replace claude with cursor, windsurf, or continue for those clients.)

Install from source

git clone https://github.com/rocnubie/wordcast-mcp.git
cd wordcast-mcp
pnpm install

Then add to your MCP client config (claude_desktop_config.json for Claude Desktop, mcp.json for Cursor / Windsurf / Continue):

{
  "mcpServers": {
    "wordcast-mcp": {
      "command": "node",
      "args": [
        "/absolute/path/to/wordcast-mcp/src/index.mjs"
      ]
    }
  }
}

Debug with MCP Inspector

npx @modelcontextprotocol/inspector node src/index.mjs

Official Links

• Website: https://wordcast.app
• Blog: https://wordcast.app/blog
• About: https://wordcast.app/about
• Support: support@wordcast.app

Development

pnpm install
pnpm start                 # run the server over stdio

License

MIT