Newgen RAG Assistant

A VS Code extension that brings local document Q&A capabilities directly into your editor. Upload PDFs or DOCX files, ask questions, and get instant answers with source citations—all running locally without any API keys.

Features

✨ Document Upload - Upload PDF and DOCX files (TXT coming soon)
🔍 Intelligent Search - Query your documents with natural language
📍 Source Citations - Get exact page, section, and chunk references
⚡ Blazing Fast - Local processing, no cloud dependencies
🔒 Privacy First - Your documents never leave your machine
🎯 Integrated Sidebar - Access directly from VS Code without context switching

Getting Started

Prerequisites

• VS Code 1.85 or later
• Python 3.8+ with FastAPI and uvicorn
• pip dependencies: fastapi, uvicorn, pypdf, python-docx, langchain

Installation

1. From VS Code Marketplace (soon):

   • Open VS Code Extensions (Ctrl+Shift+X)
   • Search for "Newgen RAG Assistant"
   • Click Install

2. From .vsix file (manual):

   • Download the latest .vsix file from releases
   • In VS Code, run: Extensions: Install from VSIX
   • Select the downloaded file

3. From Source (development):

   git clone https://github.com/your-org/newgen-rag-project
   cd vscode-extension
   npm install  # if needed
   code .
   F5  # Launch Extension Development Host
   

Usage

1. Open the Extension

   • The RAG Assistant sidebar appears automatically when you start VS Code
   • Click the ⬆ Upload button to select a document

2. Upload a Document

   • Choose a PDF or DOCX file
   • Wait for processing (status shows "READY" when done)
   • Sidebar displays: filename, page count, and chunk count

3. Ask Questions

   • Click the ❓ Ask button
   • Type your question
   • Click Send to get instant answers
   • Answers include source citations with exact locations

Architecture

┌─────────────────────┐
│   VS Code Sidebar   │
│  (HTML/CSS/JS UI)   │
└──────────┬──────────┘
           │ HTTP
           ▼
┌─────────────────────┐
│   FastAPI Backend   │
│    (port 8001)      │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│  Document Processor │
│  (PDF/DOCX Parser)  │
└──────────┬──────────┘
           │
           ▼
┌─────────────────────┐
│   Vector DB/Cache   │
│   (In-Memory)       │
└─────────────────────┘

Configuration

The extension auto-starts the FastAPI backend on port 8001. To customize:

1. Edit app.py to change ports or processing settings
2. Edit extension.js line 5 to change APP_URL

Troubleshooting

Extension won't load?

• Check VS Code version (requires 1.85+)
• Look for errors in Extension Host console (Help > Toggle Developer Tools)

Backend won't start?

• Verify Python 3.8+ is installed: python --version
• Install dependencies: pip install -r requirements.txt
• Check if port 8001 is in use: netstat -an | findstr 8001

Upload fails?

• File size limits depend on available RAM
• Supported formats: PDF, DOCX, TXT
• Check backend console for processing errors

No answers returned?

• Ensure document is fully processed (status = "READY")
• Try simpler questions first
• Check document quality (text must be extractable)

Development

Building the Extension

# Install vsce (VS Code Extension CLI)
npm install -g vsce

# Package into .vsix
cd vscode-extension
vsce package

# Result: newgen-rag-assistant-1.0.0.vsix

Publishing to VS Code Marketplace

# Create Azure DevOps account and personal access token
# Then:
vsce publish -p <YOUR_PAT>

See VS Code Extension Publishing Guide

License

MIT - See LICENSE file

Support

For issues, feature requests, or questions:

• 📧 Email: support@newgen.local
• 🐛 GitHub Issues: https://github.com/your-org/newgen-rag-project/issues
• 💬 Discussions: https://github.com/your-org/newgen-rag-project/discussions

Made with ❤️ for developers who value privacy and speed.