Networkx Mcp Server

NetworkX MCP Server

Academic-focused graph analysis in your AI conversations - The first and only NetworkX MCP server specialized for academic research and citation analysis.

[](https://github.com/Bright-L01/networkx-mcp-server/actions/workflows/ci.yml) [](https://github.com/Bright-L01/networkx-mcp-server/actions/workflows/release.yml) [](https://github.com/Bright-L01/networkx-mcp-server/actions/workflows/security.yml) [](https://github.com/Bright-L01/networkx-mcp-server/actions/workflows/docker-build.yml) [](https://pypi.org/project/networkx-mcp-server/) [](https://www.python.org/downloads/) [](https://networkx.org/) [](https://modelcontextprotocol.io/) [](https://opensource.org/licenses/MIT) [](https://crossref.org/)

🎓 What is this?

NetworkX MCP Server enables Large Language Models (like Claude) to perform sophisticated academic research and citation analysis directly within conversations. Built specifically for researchers, academics, and students who need to analyze citation networks, calculate author impact metrics, and discover literature patterns.

Stop switching between VOSviewer, CitNetExplorer, and manual analysis. Start doing academic research in your AI conversations.

🎯 Key Features

🔬 Academic Research Tools

• Citation Network Analysis: Build citation networks from DOIs using CrossRef API
• Author Impact Metrics: Calculate h-index, total citations, and academic influence
• Literature Discovery: Automated paper recommendations based on citation patterns
• Collaboration Analysis: Map co-authorship networks and identify key researchers
• Research Trend Detection: Analyze publication and citation trends over time

📊 Core Graph Operations

• 43 Graph Tools: From basic operations to advanced algorithms like PageRank
• BibTeX Export: Export citation networks in academic-standard BibTeX format
• CrossRef Integration: Access 156+ million academic papers via DOI resolution
• Visualization: Generate publication-ready network visualizations
• First of Its Kind: The only academic-focused NetworkX MCP server

🌟 Why NetworkX MCP Server for Academic Research?

• Built for Researchers: Designed specifically for academic workflows and citation analysis
• Real-time Literature Discovery: Find related papers and collaboration opportunities instantly
• Reproducible Research: Python-based, version-controlled, and shareable analysis workflows
• Academic Data Integration: Direct access to CrossRef's 156+ million paper database
• No Enterprise Complexity: Focus on research, not IT infrastructure
• Cost-Effective: Free alternative to expensive commercial citation analysis tools

📦 Installation

pip install networkx-mcp-server

🚀 Quick Start

1. Install the server

pip install networkx-mcp-server

2. Configure Claude Desktop

Add to your claude_desktop_config.json:

MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "networkx": {
      "command": "python",
      "args": ["-m", "networkx_mcp"]
    }
  }
}

3. Restart Claude Desktop

The NetworkX tools will now be available in your conversations!

🧪 Test It Works

Ask Claude: "Create a graph called 'test', add nodes 1, 2, 3 with edges between them, then find the shortest path from 1 to 3"

📊 Available Operations

🔬 Academic Research Functions

• resolve_doi - Resolve DOI to publication metadata using CrossRef API
• build_citation_network - Build citation networks from seed DOIs
• analyze_author_impact - Calculate h-index and impact metrics for authors
• find_collaboration_patterns - Analyze co-authorship networks
• detect_research_trends - Identify publication and citation trends over time
• recommend_papers - Get paper recommendations based on citation patterns
• export_bibtex - Export citation networks in BibTeX format

📊 Core Graph Operations

• create_graph - Create directed or undirected graphs
• add_nodes / remove_nodes - Add or remove nodes
• add_edges / remove_edges - Add or remove edges
• get_info - Get basic graph statistics
• list_graphs - List all stored graphs
• delete_graph - Delete a graph from storage
• shortest_path - Find optimal paths between nodes
• get_neighbors - Get all neighbors of a node
• set_node_attributes / get_node_attributes - Manage node metadata
• set_edge_attributes / get_edge_attributes - Manage edge weights and metadata

🔍 Analysis Operations

• degree_centrality - Find the most connected nodes
• betweenness_centrality - Identify bridges and key connectors
• centrality_measures - Multiple centrality metrics at once
• pagerank - Google's PageRank algorithm for node importance
• connected_components - Find isolated subgraphs
• community_detection - Discover natural groupings
• clustering_coefficients - Measure local clustering
• graph_statistics - Comprehensive graph statistics
• minimum_spanning_tree - Find minimum spanning tree
• cycles_detection - Detect cycles in a graph
• graph_coloring - Greedy vertex coloring
• matching - Maximum weight matching
• maximum_flow - Maximum flow in directed graphs
• topological_sort - Topological ordering of DAGs
• subgraph - Extract induced subgraph as new graph
• merge_graphs - Compose two graphs into one

🎨 Visualization & I/O

• visualize_graph - Create PNG visualizations with multiple layouts
• import_csv - Load graphs from edge lists
• export_json - Export graphs in standard formats

Academic Research Example

Human: Analyze citation patterns for the paper "Attention Is All You Need"

Claude: I'll help you analyze citation patterns for that influential paper.

[Resolves DOI: 10.5555/3295222.3295349]
Found paper: "Attention Is All You Need" by Vaswani et al. (2017)
Citations: 82,892 | Journal: NIPS

[Builds citation network from seed DOI]
Built citation network with 847 nodes and 2,341 edges from 2-hop analysis

[Analyzes author impact]
Ashish Vaswani: h-index 45, total citations 127,436
Most impactful paper: "Attention Is All You Need" (82,892 citations)

[Finds collaboration patterns]
Key collaborators: Noam Shazeer (Google), Niki Parmar (Google)
Research cluster: Google Brain team with 47 collaborations

[Detects research trends]
Trend: MASSIVE INCREASE in attention mechanism research post-2017
2017: 12 papers → 2023: 3,847 papers (320x growth)

[Recommends related papers]
Top recommendations based on co-citation patterns:
1. "BERT: Pre-training of Deep Bidirectional Transformers" (2018)
2. "GPT-2: Language Models are Unsupervised Multitask Learners" (2019)
3. "RoBERTa: A Robustly Optimized BERT Pretraining Approach" (2019)

[Exports BibTeX]
Generated BibTeX file with 847 entries ready for LaTeX integration

🎓 Academic Use Cases

1. Literature Review & Meta-Analysis

• Automatically expand citation networks from key papers
• Identify research gaps and emerging trends
• Calculate field-wide impact metrics
• Generate comprehensive BibTeX databases

2. Collaboration Network Analysis

• Map research collaborations within and across institutions
• Identify key researchers and potential collaborators
• Analyze interdisciplinary connections
• Study research community evolution

3. Citation Pattern Analysis

• Track knowledge diffusion through citation networks
• Identify influential papers and breakthrough research
• Analyze citation bias and self-citation patterns
• Study geographic and institutional citation patterns

4. Research Trend Detection

• Identify emerging research areas and hot topics
• Analyze publication volume and citation trends
• Track research lifecycle from emergence to maturity
• Predict future research directions

5. Academic Impact Assessment

• Calculate comprehensive author impact metrics
• Compare researchers across different career stages
• Analyze journal and conference impact patterns
• Study citation half-life and research longevity

📈 Performance

• Memory: ~70MB (including Python, NetworkX, and visualization)
• Graph Size: Tested up to 10,000 nodes
• Operations: Most complete in milliseconds
• Visualization: 1-2 seconds for complex graphs

🛠️ Development

Running from Source

# Clone the repository
git clone https://github.com/Bright-L01/networkx-mcp-server
cd networkx-mcp-server

# Install dependencies
pip install -e ".[dev]"

# Run the server
python -m networkx_mcp

Running Tests

pytest tests/working/

📚 Documentation

• [API Reference](docs/API.md) - Detailed operation descriptions
• [Contributing](CONTRIBUTING.md) - How to contribute

🤝 Contributing

We welcome contributions! This is the first NetworkX MCP server, and there's lots of room for improvement:

• Add more graph algorithms
• Improve visualization options
• Add graph file format support
• Optimize performance
• Write more examples

📄 License

MIT License - See [LICENSE](LICENSE) for details.

🙏 Acknowledgments

• NetworkX - The amazing graph library that powers this server
• Anthropic - For creating the Model Context Protocol
• The MCP community - For inspiration and examples

---

Built with ❤️ for the AI and Graph Analysis communities

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: brightlikethelight
• Source: brightlikethelight/networkx-mcp-server
• License: MIT
• Homepage: https://networkx-mcp-server.readthedocs.io

Install and usage instructions live in the source repository linked above.