MCP For Paper Read Based On Ai Ide

🚀 Research Paper Auxiliary Reading MCP System (Local Version)

This system is built on MCP (Model Context Protocol), offering local PDF parsing, structured extraction, in - depth mathematical formula analysis, code generation, and visualization capabilities. All computations are performed locally, and it supports enhancement via the built - in Trae large model or a local LLM.

✨ Features

• 📝 Smart Summarization and Methodology: The summarizer supports intelligent summarization and methodology extraction, and can automatically connect to a local LLM (Ollama) for in - depth content understanding.
• 🔢 Advanced Mathematical Intelligence: The math_explainer extracts formulas, constructs AST (Abstract Syntax Trees), and stores them in a local SQLite database.
• 💻 Experiment Reproduction Assistance: The code_generator automatically extracts hyperparameters from papers and generates PyTorch model definitions and training scripts.
• 📊 Visualization Charts: The visualization generates Mermaid flowcharts and variable dependency graphs.
• 📑 Smart Report Generation: The report_generator automatically generates a Markdown - formatted paper analysis report, including summaries, structures, visualization charts, and code configurations, while ensuring correct rendering of mathematical formulas.
• 🧠 Knowledge Management: The local database papers.db automatically stores paper metadata, symbol definitions, and experimental records.

📦 Installation

Prerequisites

• Node.js (v16+)
• Git
• (Optional) Ollama for local large - model acceleration (default port 11434)

1. Clone the Repository

git clone https://github.com/Lxy-hqu/-mcp-for-paper-read-based-on-AI-IDE.git
cd -mcp-for-paper-read-based-on-AI-IDE

2. Install Dependencies

npm install

3. Compile the Project

npx tsc

4. Configure Trae (or Cursor)

Automatic Configuration (Recommended)

This repository contains a trae_mcp_config.json template. You need to modify the command path to the absolute path of your local Node.js and args to the absolute path of server.js.

Example Configuration for Windows Users: Open Trae settings -> MCP Servers -> Edit in settings.json:

{
  "mcpServers": {
    "local-papers": {
      "command": "C:\\Program Files\\nodejs\\node.exe", 
      "args": [
        "E:\\path\\to\\-mcp-for-paper-read-based-on-AI-IDE\\dist\\server.js"
      ],
      "disabled": false,
      "autoApprove": []
    }
  }
}

💡 Usage Tip

Run where node in the terminal to get the installation path of Node.js.

5. Verify the Installation

After restarting Trae, local-papers should show as Connected in the MCP management panel. You can test it by entering in the dialog box: "Please use pdf_loader to read this file: E:\my_paper.pdf".

🚚 Migration and Multi - Machine Deployment

There are two ways to run this service on another machine:

Method 1: Direct Directory Copy (Portable)

You can directly copy the entire directory (including dist and node_modules) to the new machine.

Precautions:

1. Node.js Environment: The new machine must have Node.js installed.
2. OS Compatibility: If you copy from Windows to macOS/Linux, you may encounter errors due to incompatible native dependencies such as better - sqlite3. In this case, run npm rebuild on the new machine or delete node_modules and run npm install again.
3. Path Configuration: In Trae's settings.json, you must modify command (Node path) and args (server.js path) to the actual absolute paths on the new machine.

Method 2: Re - clone (Recommended)

Run git clone, npm install, and npx tsc on each new machine. This is the most reliable method to ensure all dependencies are perfectly compatible with the current system.

⚠️ Important Note

The papers.db file stores parsed paper data. Directly copying the directory will transfer this database. If you need a clean environment on the new machine, delete the papers.db file.

💻 Usage Examples

Mathematical Formula Analysis

"Analyze the mathematical formulas in this paper and store the symbol definitions in the database."

Code Generation

"Generate PyTorch model code based on the method section of this paper and extract hyperparameters."

Visualization

"Generate a Mermaid flowchart for the model structure of this paper."

Generate Analysis Report

"Generate a complete Markdown analysis report for this paper."

📁 Project Structure

src/
├── providers/          # Core functional modules
│   ├── pdf_loader.ts      # PDF reading
│   ├── structure_parser.ts # Structured segmentation
│   ├── math_explainer.ts   # Mathematical analysis & AST
│   ├── code_generator.ts   # Code & Config generation
│   ├── visualization.ts    # Mermaid chart generation
│   ├── report_generator.ts # Report generation
│   └── summarizer.ts       # Smart summarization and methodology
├── database.ts         # SQLite database management
├── llm_client.ts       # Local LLM connection client
├── server.ts           # MCP server entry
└── types.d.ts          # Type definitions

📄 License

MIT