Explore the demo paper’s supporting assets in the /demo folder.

PaperDebugger is an AI-powered academic writing assistant that helps researchers debug and improve their research papers with intelligent suggestions and seamless Overleaf integration, without leaving the editor. It is powered by a custom MCP-based orchestration engine that simulates the full academic workflow Research → Critique → Revision. This enables multi-step reasoning, reviewer-style critique, and structured revision passes beyond standard chat-based assistance.

🌐 Social Media & Community Groups

Our team is actively working to improve long-term reliability, hoping to iron out issues this month. Thank you for your patience. Stay connected with the PaperDebugger community! Join our Discord or WeChat channels for updates, announcements, and support.

📋 Table of Contents

• 📋 Table of Contents

• ✨ Features

• 🎯 Quick Start

• For Users

• Custom Endpoint Configuration

• 🏗️ Architecture Overview

• 🛠️ Development Setup

• Prerequisites

• System Requirements

• Development Tools

• Quick Installation (macOS/Linux with Homebrew)

• Backend Build

• 1. Clone the Repository

• 2. Start MongoDB

• 3. Environment Configuration

• 4. Custom MCP Backend Orchestration

• 5. Build and Run

• Frontend Extension Build

• Chrome Extension Development

• Installing the Development Extension

✨ Features

PaperDebugger never modifies your project, it only reads and provides suggestions.

• 🤖 AI-Powered Chat: Intelligent conversations about your Overleaf project
• ⚡ Instant Insert: One-click insertion of AI responses into your project
• 💬 Comment System: Automatically generate and insert comments into your project
• 📚 Prompt Library: Custom prompt templates for different use cases
• 🔒 Privacy First: Your content stays secure - we only read, never modify
• 🧠 Multi-Agent Orchestration – XtraMCP support for literature-grounded research, AI-Conference review, and domain-specific revision preview.mp4

🎯 Quick Start

For Users

Install the Extension

• Chrome Web Store (Recommended)
• Latest Release (Manual Install)

Manual Installation

• Download the latest release
• Open Chrome and go to chrome://extensions/
• Enable "Developer mode"
• Click "Load unpacked" or drag the extension file

Start Using

• Open any Overleaf project
• Click the PaperDebugger icon
• Begin chatting with your LaTeX assistant!

Custom Endpoint Configuration

If you want to use a self-hosted PaperDebugger backend, you can configure a custom endpoint. Note: You need to handle HTTPS serving yourself, as Chrome blocks HTTP requests from HTTPS websites for security reasons.

Steps:

1. Open the PaperDebugger extension
2. Go to Settings, click the version number 5 times to enable "Developer Tools" (a.)
3. Enter your backend URL in the "Backend Endpoint" field (b.)
4. Refresh the page

If you encounter endpoint errors after refresh, use the "Advanced Options" at the bottom of the login page to reconfigure.

🏗️ Architecture Overview

The PaperDebugger backend is built with modern technologies:

• Language: Go 1.24+
• Framework: Gin (HTTP) + gRPC (API)
• Database: MongoDB
• AI Integration: OpenAI API
• Architecture: Microservices with Protocol Buffers
• Authentication: JWT-based with OAuth support

🛠️ Development Setup

Prerequisites

System Requirements

• Go: 1.24 or higher
• Node.js: LTS version (for frontend build)
• MongoDB: 4.4 or higher
• Git: For cloning the repository

Development Tools

• Buf: Protocol Buffer compiler
• Wire: Dependency injection code generator
• Make: Build automation

Quick Installation (macOS/Linux with Homebrew)

# Install Go
brew install go

# Install Buf (required for Protocol Buffers)
brew install bufbuild/buf/buf

# Install Node.js
brew install node

Backend Build

1. Clone the Repository

git clone https://github.com/PaperDebugger/paperdebugger.git
cd paperdebugger

2. Start MongoDB

# Using Docker (recommended)
docker run -d --name mongodb -p 27017:27017 mongo:latest

3. Environment Configuration

cp .env.example .env
# Edit the .env file based on your configuration

4. Custom MCP Backend Orchestration [OPTIONAL FOR LOCAL DEV]

Our enhanced orchestration backend, XtraMCP, is partially deployed in-production, with selected components enabled to balance stability and operational cost at this stage. It is still under active development and remains closed-source for now.

To understand the backend orchestration logic and view the currently available and gated features, as well as planned capabilities, refer to /demo/xtramcp/readme.md. You can still self-host PaperDebugger without it; all core features (chat, formatting, edits, comments) work normally.

Connecting to XtraMCP unlocks:

• research-mode agents (with on-demand search for latest publications),
• structured reviewer-style critique,
• domain-specific revisions tailored for academic writing powered by XtraGPT models.

We plan to open-source XtraMCP once the API stabilizes for community use.

5. Build and Run

# Build the backend
make build

# Run the backend server
./dist/pd.exe

The server will start on http://localhost:6060.

NOTE: "ERROR [AI Client] Failed to initialize XtraMCP session" is expected if you’re hosting locally without XtraMCP or an equivalent MCP orchestration backend.

Frontend Extension Build

Chrome Extension Development

cd webapp/_webapp

# Install frontend dependencies
npm install

# Build for production (connects to production server)
npm run build:prd:chrome

# Package the extension
cd dist
zip -r paperdebugger-extension.zip *

Installing the Development Extension

1. Open Chrome and navigate to chrome://extensions/
2. Enable "Developer mode" (toggle in top-right)
3. Click "Load unpacked" and select the webapp/_webapp/dist directory

• Or drag the paperdebugger-extension.zip file into the extensions page

Citation

@misc{hou2025paperdebugger,
title={PaperDebugger: A Plugin-Based Multi-Agent System for In-Editor Academic Writing, Review, and Editing},
author={Junyi Hou and Andre Lin Huikai and Nuo Chen and Yiwei Gong and Bingsheng He},
year={2025},
eprint={2512.02589},
archivePrefix={arXiv},
primaryClass={cs.AI},
url={https://arxiv.org/abs/2512.02589},
}