Skei

High-performance interactive visualization and query interface for large Neo4j graphs.

Neo4j is usually the bottleneck with layouts being completed in under 3 seconds even for graphs larger than 10k nodes and 15k edges.

Scope

This repository is focused on Neo4j graph exploration.

• Run Cypher queries and render the result as both graph and table views.
• Precompute layouts server-side so larger graph results stay navigable.
• Explore results with zoom, pan, search, highlighting, and minimap support.

Examples

UI Overview

Example Graph Outputs

Quick Start

1. Create a virtual environment and install dependencies.

python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt

python -m venv .venv
source .venv\bin\activate
pip install -r requirements.txt

2. Set Neo4j connection details if needed.

set NEO4J_HTTP_URL=http://localhost:7474
set NEO4J_USER=neo4j
set NEO4J_PASSWORD=your-password
set NEO4J_DATABASE=neo4j

3. Run the app.

python app.py

4. Open http://localhost:5555.

You can also change the connection at runtime from the sidebar in the UI.

Features

• Neo4j HTTP Transactional API integration.
• Automatic layout selection for graph size.
• Canvas-based rendering with D3 zoom and pan.
• Curved edges, neighborhood highlighting, node search, and minimap.
• Graph view and table view for the same query result.
• Demo graph generation for UI and performance testing without a live Neo4j instance.

Layouts

Supported layout modes:

• auto
• force_directed
• force_directed_hq
• community
• circle
• drl
• kamada_kawai
• spectral

Availability depends on whether python-igraph is installed.

API Endpoints

• GET /: Explorer UI.
• POST /api/query: Execute Cypher and return graph data, table rows, and timings.
• POST /api/relayout: Recompute layout for the cached query result.
• GET /api/schema: Return labels, relationship types, and property keys.
• GET /api/connection-test: Verify Neo4j connectivity.
• POST /api/reconnect: Update Neo4j connection settings at runtime.
• GET /api/layouts: Return available layout algorithms.
• GET /api/sample-queries: Return built-in sample Cypher queries.
• POST /api/demo: Generate synthetic graph data for demo and performance testing.

Configuration

Environment variables used by app.py:

• NEO4J_HTTP_URL default: http://localhost
• NEO4J_USER default: neo4j
• NEO4J_PASSWORD default: empty
• NEO4J_DATABASE default: neo4j
• SAMPLE_QUERIES_FILE default: config/sample_queries.json

Project Structure

• app.py: Flask app and API endpoints.
• layout_engine.py: Graph layout computation and algorithm selection.
• templates/index.html: Frontend application.
• config/sample_queries.json: Built-in sample Cypher queries.
• requirements.txt: Python dependencies.

Notes

• The supported service entry point is python app.py.
• The repository is intentionally scoped to Neo4j graph visualization and layout exploration.