How mwasifanwar/cryptograph works
Project type and purpose
• From the README and file tree, this is a full-stack, data-driven analytics platform named “cryptograph.”
• Its core function is to ingest raw blockchain data, construct transaction graphs, extract features, run a graph neural network (GNN) and anomaly-detection models, and present results via a web interface (templates + static assets).
File structure
• Top-level modules:
• data: blockchain_loader.py, graph_builder.py – responsible for fetching and structuring raw data.
• features: feature_engineer.py – computing node/edge features for GNN input.
• analysis: pattern_analyzer.py, risk_assessor.py – orchestrating detection logic on model outputs.
• models: gnn_models.py, anomaly_detector.py – definition and execution of machine-learning models.
• api: app.py – Flask (or similar) app exposing REST endpoints for the front end.
• static + templates: UI layer with HTML, CSS, JS.
• config: settings.py – configuration of endpoints, thresholds, feature toggles.
• utils: helper functions (blockchain_utils.py, visualization_utils.py).
• main.py – likely CLI entrypoint or pipeline orchestrator.
README
• Confirms GNN-based money-laundering/fraud detection in blockchain transactions and smart contracts.