Summary

PaperDebugger is an AI-powered academic writing assistant for Overleaf, designed to debug and enhance LaTeX papers. It offers intelligent suggestions, seamless integration, and advanced features like multi-step reasoning and reviewer-style critique to streamline the academic writing and revision process.

How It Works

The system utilizes a custom MCP-based orchestration engine simulating an academic workflow (Research → Critique → Revision). This enables multi-step reasoning and structured revision passes beyond standard chat assistants. A core principle is that PaperDebugger only reads project files, ensuring user content remains secure and unmodified.

Quick Start & Requirements

Installation is via the Chrome Web Store or manual download of the latest release. Users enable developer mode in Chrome and load the unpacked extension. Development prerequisites include Go 1.24+, Node.js LTS, MongoDB 4.4+, Git, Buf, and Make. Backend setup involves cloning the repository, starting MongoDB (Docker recommended), configuring .env, building with make build, and running the server. Frontend development requires npm install and npm run build:prd:chrome. Custom backend endpoints require self-hosted HTTPS.

Highlighted Details

• AI-Powered Chat: Intelligent conversations directly within the Overleaf editor.
• Instant Insert & Comment System: Allows one-click insertion of AI responses and automatic generation/insertion of comments.
• Prompt Library: Provides custom prompt templates for various academic tasks.
• Privacy First: Guarantees that user project content is only read, never modified.
• XtraMCP Orchestration: Supports advanced multi-agent capabilities for literature-grounded research, AI-driven conference reviews, and domain-specific revisions, though this component is currently closed-source.

Maintenance & Community

The README provides no specific links to community channels (e.g., Discord, Slack), a public roadmap, or details on notable contributors or sponsorships.

Licensing & Compatibility

The project's license is not explicitly stated in the provided README.

Limitations & Caveats

Advanced XtraMCP features (research agents, reviewer critique) are closed-source and under active development, with future open-sourcing planned. Core functionalities are available without this component. Self-hosting custom backends requires managing HTTPS serving.