Autonomous Scientist Desktop Extension v6.0

A comprehensive research automation system for social sciences and humanities, built as a Claude Desktop Extension (DXT).

🙏 Acknowledgments

This project was inspired by and builds upon the excellent work from:

• AI-Scientist v2 - Advanced automated scientific research capabilities
• Lotus Wisdom MCP - Model Context Protocol implementation patterns

Special thanks to these projects for demonstrating the potential of AI-assisted research automation.

🎯 Overview

The Autonomous Scientist is an advanced research assistant that combines:

• Multi-language OCR processing for academic PDFs
• Discipline-specific analysis for 8 academic fields
• LaTeX document generation with proper citation formatting
• Comprehensive literature search across multiple academic databases
• Hardware optimization for Intel i3-12100F + 16GB DDR4

🔬 Supported Disciplines

1. Psychology - APA 7th Edition formatting, experimental design analysis
2. Neuroscience - Nature format, neuroimaging methodology analysis
3. Education - APA Educational format, curriculum and pedagogy analysis
4. Sociology - ASA format, social network and demographic analysis
5. Anthropology - AAA format, ethnographic and cultural analysis
6. Philosophy - Chicago 17th Edition, argument structure analysis
7. Political Science - APSA format, institutional and electoral analysis
8. International Relations - APSA format, conflict and diplomacy analysis

📦 Installation as Desktop Extension

Prerequisites

Install the DXT CLI:

npm install -g @anthropic-ai/dxt

Build and Install

1. Build the extension:

   cd autonomous-scientist-extension
   npm install
   npm run build
   npm run pack

2. Install in Claude Desktop:

   • Open Claude Desktop
   • Navigate to Extensions → Install Extension
   • Select the generated .dxt file
   • Follow the configuration wizard

3. Configure settings:

   • Set your primary research discipline
   • Configure citation style preferences
   • Set workspace directory
   • Add API keys (optional but recommended)

🔧 Configuration Options

The extension supports extensive user configuration through Claude Desktop's settings:

Required Settings

• Primary Research Discipline: Your main field of study
• Default Citation Style: APA, Chicago, MLA, ASA, AAA, or APSA
• Workspace Directory: Where to store research projects

Optional Settings

• API Keys: Semantic Scholar, CrossRef, OpenAI, Anthropic
• Cache Size: Disk cache limit (1-20GB)
• OCR Languages: Comma-separated language codes
• Advanced Features: Enable experimental functionality

API Integration

• Semantic Scholar: 200M+ academic papers (free)
• ArXiv: Latest preprints (free)
• CrossRef: Publisher metadata (free)
• OpenAI: Enhanced analysis (paid, optional)
• Anthropic: Extended Claude access (paid, optional)

🛠️ Core Tools (40+ Available)

Literature & Research

• comprehensive_literature_search - Multi-source academic search
• analyze_by_discipline - Specialized analysis for your field
• identify_research_gaps - Automatic gap identification

PDF Processing

• process_academic_pdf - Complete PDF analysis with OCR
• batch_process_pdfs - Efficient batch processing
• ocr_multilingual - Advanced multi-language OCR

Document Generation

• generate_latex_paper - Complete LaTeX documents
• format_citations - Multi-style citation formatting
• compile_to_pdf - LaTeX compilation with error handling

Discipline-Specific Analysis

• analyze_psychology_research - APA methodology analysis
• analyze_neuroscience_paper - Neuroimaging analysis
• analyze_education_study - Pedagogical analysis
• analyze_philosophy_argument - Argument structure
• And 4 more specialized analyzers...

📚 Usage Examples

Literature Review Workflow

User: "I need a literature review on cognitive behavioral therapy"
→ Extension searches multiple databases
→ Processes provided PDFs with OCR
→ Analyzes content by psychology discipline
→ Generates complete LaTeX literature review

Document Analysis

User: "Analyze this neuroscience paper" [attach PDF]
→ OCR processing with quality enhancement
→ Extracts methodology, brain regions, findings
→ Provides quality assessment and recommendations
→ Formats citations in desired style

Multilingual Research

User: Uploads German psychology paper (scanned)
→ Auto-detects German language
→ Applies academic OCR enhancement
→ Analyzes with psychology-specific patterns
→ Generates English summary with APA citations

🚀 Predefined Workflows

Literature Review Workflow

Comprehensive workflow for conducting literature reviews:

1. Multi-source literature search
2. PDF processing and analysis
3. Gap identification
4. LaTeX document generation

Document Analysis Workflow

Deep analysis of academic documents:

1. OCR processing with quality enhancement
2. Discipline-specific content analysis
3. Citation extraction and formatting
4. Research recommendations

💻 System Requirements

Minimum Requirements

• CPU: Intel i3-12100F or equivalent (4 cores)
• RAM: 16GB DDR4
• Storage: 5GB free space for cache
• OS: Windows 10/11, macOS 12+, or Linux
• Claude Desktop: v0.8.0 or later
• Node.js: v18.0.0 or later

Performance Optimization

• Memory usage optimized for 16GB systems
• Intelligent caching (2GB memory + 5GB disk)
• Adaptive processing based on available resources
• Hardware-aware concurrency limits

🔒 Security & Privacy

• Local Processing: All PDF processing occurs on your machine
• Encrypted Storage: API keys stored with AES-256 encryption
• No Data Sharing: Research content never leaves your system
• Secure APIs: Only established academic APIs
• Optional Cloud: Premium APIs are optional enhancements

🐛 Troubleshooting

Common Issues

High Memory Usage

• Extension automatically manages memory limits
• Reduce concurrent PDF processing if needed
• Clear cache through extension settings

OCR Quality Issues

• Ensure PDF resolution >150 DPI
• Try different language settings
• Use quality enhancement for scanned documents

API Connection Failures

• Check internet connection
• Verify API keys in settings
• Test individual API connections

🔄 Development

Current Status (v6.0)

✅ Stable Components:

• Simplified MCP Server (server/index-simple.js) - Core functionality working
• DXT Package Integration - Properly formatted for Claude Desktop
• Basic Research Tools - Literature search, PDF processing, API setup

🚧 In Development:

• Full TypeScript Conversion - Complex tools being migrated from TypeScript
• Advanced PDF Processing - OCR and citation extraction
• Discipline-Specific Analyzers - Specialized analysis tools

Recent Fixes (January 2025)

Issue Resolution:

• ✅ Server Disconnection: Fixed TypeScript compatibility issues by creating simplified server
• ✅ MCP JSON Parsing: Resolved stdout pollution in mcp-science-web wrapper
• ✅ Package Installation: Updated manifest.json for proper DXT structure

Architecture Decision: The project now uses a hybrid approach:

1. Simplified Server (index-simple.js) for immediate functionality
2. Full Implementation (index.js) for advanced features (in progress)

This ensures users have a working research assistant while development continues.

Dependencies

Runtime Requirements:

• Node.js 18+ with ES Module support
• Python 3.8+ with MCP SDK
• Git for version control

Optional Integrations:

• LaTeX distribution for document generation
• Tesseract OCR for multilingual text extraction

Project Structure

autonomous-scientist-extension/
├── manifest.json          # DXT manifest (updated for simplified server)
├── server/                # MCP server implementation
│   ├── index-simple.js    # Simplified working server (current)
│   ├── index.js           # Full server (TypeScript conversion in progress)
│   ├── tools/             # Tool implementations (40+ tools)
│   └── utils/             # Utility modules (security, memory, cache)
├── assets/                # Icons and screenshots
├── templates/             # LaTeX templates for disciplines
└── autonomous-scientist-dxt.dxt  # Packaged extension file

Building from Source

1. Clone the repository
2. Install dependencies: npm install
3. Build TypeScript: npm run build
4. Package DXT: npm run pack

Contributing

• Follow the existing code structure
• Implement proper error handling
• Add comprehensive logging
• Test with multiple document types

📄 License

MIT License - See LICENSE file for details

🙋‍♂️ Support

For issues and support:

• Check the troubleshooting section
• Review Claude Desktop Extension documentation
• Report bugs with system specifications
• Include example files (anonymized) for OCR issues

---

Transform your research workflow with the Autonomous Scientist Desktop Extension - the most comprehensive academic research automation tool for social sciences and humanities.

Optimized for your hardware • Secure by design • Research-focused