Home / Servers

mcp-nano

by ChrisMares

mcp-nano

Native desktop rewrite of VectorFlow, built with Tauri 2, Rust, React, Vite, and Tailwind CSS.

The full backend is implemented in Rust: ONNX Runtime ML models (dense embedding, BM25 sparse, cross-encoder reranker), tree-sitter code chunkers (11 languages), document loaders (PDF, DOCX, XLSX, HTML, CSV, XML, ODT, Markdown, plain text), SQLite persistence, Qdrant vector DB sidecar management, hybrid RAG query pipeline, MCP streamable-HTTP server on port 18651, background job worker, and website crawling/embedding — all running locally with zero Docker or Python dependencies.

Prerequisites

  • Node.js 22 or later
  • npm 10 or later
  • Rust stable via rustup
  • Linux only: Tauri system dependencies

On Ubuntu 24.04, install the Linux dependencies:

sudo apt update
sudo apt install -y \
  build-essential curl wget file patchelf pkg-config libxdo-dev \
  libwebkit2gtk-4.1-dev libssl-dev libgtk-3-dev \
  libayatana-appindicator3-dev librsvg2-dev libdbus-1-dev

Install

npm install

Run The Frontend

Starts Vite at http://localhost:18674. This port is intentionally separate from the original VectorFlow project.

npm run dev

Run The Desktop App

Starts Vite and opens the application in a native Tauri window:

npm run tauri dev

Build

Prerequisites (one-time)

Ensure model files and the Qdrant binary are downloaded:

bash scripts/download-models.sh
bash scripts/download-qdrant.sh

Linux (Ubuntu/Debian)

# Install system dependencies
sudo apt update
sudo apt install -y build-essential curl wget file patchelf pkg-config \
  libxdo-dev libwebkit2gtk-4.1-dev libssl-dev libgtk-3-dev \
  libayatana-appindicator3-dev librsvg2-dev libdbus-1-dev

# Build the .deb installer
npm install
npm run tauri build

# Output: src-tauri/target/release/bundle/deb/mcp-nano_*.deb

For an .AppImage instead of .deb, change "targets" in tauri.conf.json under bundle to ["appimage"].

Windows

# 1. Install Rust via rustup.msi (https://rustup.rs)
# 2. Install Node.js 22+ (https://nodejs.org)
# 3. Install Visual Studio Build Tools or VS 2022 with "Desktop development with C++"
# 4. WebView2 is included on Windows 10+ / Microsoft Edge

# Build the .msi installer
npm install
npm run tauri build

# Output: src-tauri/target/release/bundle/msi/mcp-nano_*.msi

Frontend-only build

Build the production frontend bundle (no native app):

npm run build

Output is written to dist/. Tauri build artifacts are under src-tauri/target/release/.

Project Layout

src/                 React application source
src-tauri/           Rust and Tauri application
vite.config.ts       Root Vite configuration for the ported frontend

Useful Commands

npm run lint         # Run ESLint against the frontend
npm run test         # Run frontend tests
npm run preview      # Serve the production frontend bundle locally

Related servers

n8n

by n8n-io

Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.

NOASSERTION198,904

@modelcontextprotocol/server-everything

Official

by modelcontextprotocol

MCP server that exercises all the features of the MCP protocol

89,105

@modelcontextprotocol/server-filesystem

Official

by modelcontextprotocol

MCP server for filesystem access

SEE LICENSE IN LICENSE89,105

mcp-server-fetch

Official

by modelcontextprotocol

A Model Context Protocol server providing tools to fetch and convert web content for usage by LLMs

89,105