Cognee Visualization

Interactive HTML knowledge-graph visualization for Cognee-Rust.

This crate ports the Python cognee_network_visualization module. It reads all nodes and edges from any GraphDBTrait implementation and renders them into a single self-contained HTML file that uses d3.js v7 for force-directed layout and Canvas rendering.

Usage

use cognee_graph::GraphDBTrait;
use cognee_visualization::visualize;
use std::path::Path;

async fn example(graph_db: &dyn GraphDBTrait) -> Result<(), Box<dyn std::error::Error>> {
    // Write the visualization to a caller-specified file.
    let path = visualize(graph_db, Some(Path::new("/tmp/graph.html"))).await?;

    // Or write to ~/graph_visualization.html (matches Python behavior).
    let path = visualize(graph_db, None).await?;

    println!("wrote {}", path.display());
    Ok(())
}

API

• visualize(graph_db, output_path) -> PathBuf — render and write the HTML file. When output_path is None, writes to ~/graph_visualization.html (%USERPROFILE% on Windows). Returns the path written.
• render(graph_db) -> String — render the HTML string without writing it (useful for streaming over HTTP or embedding in a larger page).
• render_multi_user(pairs) -> String — aggregate multiple (user_label, Arc<dyn GraphDBTrait>) pairs into one HTML document. Nodes are deduplicated by stringified id (first-write-wins) and tagged with a source_user attribute so the d3 template can color-code by user. Mirrors Python's aggregate_multi_user_graphs().

Errors are surfaced via VisualizationError.