Omniclip RAG

v0.4.8Knowledge & Memorystable

Local-first RAG desktop app & official MCP Server. Let any AI instantly search your private Markdown, PDF, and 1290+ document formats. (本地优先的 RAG 桌面端与官方 MCP 服务器。让任意 AI 瞬间检索你的私有 Markdown、PDF 及 1290+ 种文档格式。)

knowledge-baselocal-ragmarkdownmcp-servermodel-context-protocol
Share:
37
Stars
0
Downloads
0
Weekly
0/5

What is Omniclip RAG?

Omniclip RAG is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to local-first rag desktop app & official mcp server. let any ai instantly search your private markdown, pdf, and 1290+ document formats. (本地优先的 rag 桌面端与官方 mcp 服务器。让任意 ai 瞬间检索你的私有 markdown、pdf 及 1290+ 种文...

Local-first RAG desktop app & official MCP Server. Let any AI instantly search your private Markdown, PDF, and 1290+ document formats. (本地优先的 RAG 桌面端与官方 MCP 服务器。让任意 AI 瞬间检索你的私有 Markdown、PDF 及 1290+ 种文档格式。)

This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Local-first RAG desktop app & official MCP Server. Let any A

Use Cases

Search private Markdown, PDFs, and 1290+ document formats locally. Instantly retrieve information from your personal knowledge base.
msjsc001

Maintainer

LicenseMIT
Languagepython
Versionv0.4.8
UpdatedMay 20, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx omniclip-rag

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Omniclip RAG

OmniClip RAG is a local-first Retrieval-Augmented Generation desktop application and MCP server that indexes your private Markdown files, PDFs, and over 1,290 additional document formats using hybrid full-text search (SQLite FTS5) and vector embeddings (LanceDB with the BAAI/bge-m3 model). Once indexed, any MCP-compatible AI client can instantly search your personal knowledge base via the omniclip.search tool without uploading any documents to the cloud. Obsidian users, researchers, and knowledge workers use it to make their entire local note vault instantly queryable from within Claude or other AI clients.

Prerequisites

  • Windows 10/11 (the current release ships as a Windows executable; macOS/Linux builds may be available in later releases)
  • No API keys or cloud accounts required — fully local-first
  • The OmniClip RAG desktop app downloaded from the GitHub Releases page
  • An MCP-compatible client such as Claude Desktop, Jan.ai, or OpenClaw
1

Download and extract the desktop app

Go to the GitHub Releases page at github.com/msjsc001/OmniClip-RAG/releases and download OmniClipRAG-v0.4.8-win64.zip. Extract it and run the executable — no Python, Docker, or environment setup needed.

2

Select your vault and build the index

On first launch, select your Markdown vault root folder (e.g. your Obsidian vault). The app runs a space/time precheck, downloads the BAAI/bge-m3 embedding model once, then triggers a Full Build to index all documents.

3

Download the MCP server executable

From the same Releases page, download OmniClipRAG-MCP-v0.4.8-win64.zip. Extract it to a permanent location — the MCP server binary must remain at a stable path since you will reference it in your client config.

4

Add the MCP server to your client config

Edit claude_desktop_config.json (or your client's equivalent) to point at the extracted OmniClipRAG-MCP.exe. The server uses stdio transport and requires no arguments.

{
  "mcpServers": {
    "omniclip-rag": {
      "command": "C:\\Apps\\OmniClip RAG\\OmniClipRAG-MCP.exe"
    }
  }
}
5

Restart your MCP client and run a test search

Restart Claude Desktop (or your chosen client) and ask it to search your knowledge base. The server exposes two tools: omniclip.status (checks readiness) and omniclip.search (semantic + full-text retrieval).

Omniclip RAG Examples

Client configuration

Claude Desktop config using the OmniClipRAG-MCP.exe binary. Adjust the path to match where you extracted the MCP server zip.

{
  "mcpServers": {
    "omniclip-rag": {
      "command": "C:\\Apps\\OmniClip RAG\\dist\\OmniClipRAG-MCP-v0.4.8\\OmniClipRAG-MCP.exe",
      "args": []
    }
  }
}

Prompts to try

Example prompts that use the omniclip.search tool to retrieve information from your local knowledge base.

- "Search my local knowledge base for notes about 'attention mechanism' and summarize what I've written"
- "Use OmniClip to find all my notes related to project planning and list the key points"
- "Search OmniClip for 'privacy architecture' and then explain the tradeoffs based on my notes"
- "Is OmniClip ready? Check its status and then search for 'machine learning'"
- "Find notes about 'Obsidian plugins' in my vault and give me a summary"

Troubleshooting Omniclip RAG

omniclip.search returns no results even though notes exist

The index must be fully built before searching. Open the OmniClip RAG desktop app and check that the 'Full Build' completed successfully. If not, trigger it manually from the app's main screen. The desktop app must be running for the MCP server to access the index.

MCP server binary fails to start on Windows

Windows Defender or another antivirus may block unsigned executables. Right-click OmniClipRAG-MCP.exe, select Properties, and click 'Unblock' at the bottom of the General tab. Alternatively add the extraction folder to your antivirus exclusion list.

Embedding model download hangs or fails during first setup

The BAAI/bge-m3 model (~600MB) is downloaded from Hugging Face on first run. If you are behind a proxy or firewall, ensure outbound HTTPS to huggingface.co is allowed. The model is cached in %APPDATA%\OmniClip RAG after the first successful download.

Frequently Asked Questions about Omniclip RAG

What is Omniclip RAG?

Omniclip RAG is a Model Context Protocol (MCP) server that local-first rag desktop app & official mcp server. let any ai instantly search your private markdown, pdf, and 1290+ document formats. (本地优先的 rag 桌面端与官方 mcp 服务器。让任意 ai 瞬间检索你的私有 markdown、pdf 及 1290+ 种文档格式。) It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Omniclip RAG?

Follow the installation instructions on the Omniclip RAG GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Omniclip RAG?

Omniclip RAG works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Omniclip RAG free to use?

Yes, Omniclip RAG is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Knowledge & Memory MCP Servers

Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "omniclip-rag": { "command": "npx", "args": ["-y", "omniclip-rag"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Omniclip RAG?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides