Claude Code Plugins

Community-maintained marketplace

Feedback

knowledge-graph-builder

@jennifer-mckinney/my-skills
1
0

Build large-scale knowledge graphs with D3.js visualization, CI/CD pipelines, ETL processes, and query optimization. Supports Neo4j and graph databases for millions of nodes with incremental updates and interactive exploration.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name knowledge-graph-builder
description Build large-scale knowledge graphs with D3.js visualization, CI/CD pipelines, ETL processes, and query optimization. Supports Neo4j and graph databases for millions of nodes with incremental updates and interactive exploration.

Knowledge Graph Builder Skill

Build and visualize large-scale knowledge graphs with millions of nodes, ETL pipelines, and CI/CD automation.

What This Skill Provides

Core Tools

  • build_knowledge_graph.py - Construct graphs from data sources (JSON, CSV, databases)
  • visualize_graph.py - D3.js interactive visualizations
  • optimize_graph_queries.py - Query optimization for large graphs
  • setup_graph_pipeline.py - CI/CD pipeline for graph updates

References

  • large_scale_graphs.md - Handling 10M+ nodes, partitioning strategies
  • d3_graph_viz.md - Interactive D3.js visualizations
  • graph_etl_patterns.md - ETL pipelines for graph construction
  • best_practices.md - Graph database best practices
  • troubleshooting.md - Common graph issues

Templates

  • Graph construction pipeline templates
  • D3.js visualization components
  • GitHub Actions workflows for graph updates

When to Use

Perfect For

  • Building knowledge bases with millions of entities
  • Network analysis and relationship mapping
  • D3.js interactive graph visualization
  • CI/CD for automated graph updates
  • Graph ETL pipelines
  • Query optimization for large graphs

Not For

  • Simple graph schema design (use neo4j-integration)
  • Small graphs (< 10,000 nodes) - over-engineered
  • Relational database modeling

Quick Start

# Build graph from data
python scripts/build_knowledge_graph.py \
  --input data.json \
  --output graph.db

# Visualize
python scripts/visualize_graph.py \
  --graph graph.db \
  --output visualization.html

# Setup CI/CD
python scripts/setup_graph_pipeline.py \
  --repo-path . \
  --schedule "0 0 * * *"

Decision Trees

Which visualization approach?

  • Small (< 1000 nodes): Full D3.js force-directed
  • Medium (1K-100K): Clustered view with drill-down
  • Large (> 100K): Heatmaps, aggregated views

Quality Checklist

  • Graph schema documented
  • ETL pipeline automated
  • Incremental updates implemented
  • Indexes on frequently queried properties
  • CI/CD pipeline configured
  • Visualization responsive and interactive

Common Pitfalls

Pitfall: Full Graph Rendering

Solution: Use clustering, viewport-based rendering, or aggregation for large graphs.

Pitfall: Slow Queries

Solution: Create indexes, use query profiling, optimize Cypher/Gremlin queries.

Pro Tips

Tip 1: Incremental Updates

def update_graph_incrementally(new_data):
    # Only update changed nodes/edges
    for entity in new_data:
        if entity.hash != stored_hash:
            update_node(entity)

Related Skills

  • neo4j-integration - Schema design and basic queries
  • data-viz-studio - General data visualization
  • frontend-component-system - D3.js component patterns