How codestartai-009/capstone works
This repository provides phishing-URL risk assessment through a Flask API and web scanner, scan history, and a Chrome extension, alongside a reproducible ML pipeline. In the live workflow, a user or extension submits a URL; Flask routes it to prediction services, which invoke the model pipeline and return a risk signal. The service can add optional threat-intelligence results, and scans can be saved for history and dashboard use. The graph also shows model training and evaluation. Some frontend and route internals were not excerpted, so those links rely on README statements or are omitted.
Main entry points: routes API Routes · dataset.py Dataset Preparation · prediction_service.py Prediction Service · feature_extractor.py Feature Extraction
Components 22
Application
- Flask App
backend/__init__.py - API Routes
backend/routes
Scanner Clients
- Web Scanner
static/js/scanner.js - History Dashboard
static/js/dashboard.js - Chrome Extension
extension/popup.js
Risk Assessment
- Prediction Service
services/prediction_service.py - Threat Intelligence
services/threat_intelligence.py - Model Predictor
ml/predictor.py - URL Handling
ml/url_utils.py - Feature Extraction
ml/feature_extractor.py - Risk Explanations
services/explanation_engine.py
Scan History
- History Service
services/history_service.py - Scan Repository
database/repository.py - SQLite History
database/database.py
ML Pipeline
- Feature Schema
ml/feature_schema.py - Dataset Preparation
ml/dataset.py - Model Training
ml/train_model.py - Model Utilities
ml/model_utils.py - Model Evaluation
ml/evaluate_model.py - Error Analysis
ml/error_analysis.py
Other components
- User
- Extension User
Connections 27
- User → Web Scanner: scans URL
README.md - Extension User → Chrome Extension: checks current tab
README.md - Web Scanner → API Routes: submits scan
README.md - Chrome Extension → API Routes: requests scan
README.md - API Routes → Flask App: served by
backend/__init__.py - Flask App → API Routes: registers routes
backend/__init__.py - API Routes → Prediction Service: requests prediction
backend/routes/prediction.py - API Routes → History Service: requests history
backend/routes/history.py - Prediction Service → Model Predictor: scores URL
services/prediction_service.py - Prediction Service → Threat Intelligence: optional lookup
services/prediction_service.py - Model Predictor → Feature Extraction: extracts features
ml/predictor.py - Model Predictor → Risk Explanations: builds explanations
ml/predictor.py - Feature Extraction → URL Handling: normalizes URL
ml/feature_extractor.py - Feature Extraction → Feature Schema: uses feature definitions
ml/feature_extractor.py - History Service → Scan Repository: reads and writes scans
services/history_service.py - Scan Repository → SQLite History: persists scans
- SQLite History → Feature Schema: defines migrations
database/database.py - Dataset Preparation → Feature Extraction: recomputes features
ml/dataset.py - Dataset Preparation → Feature Schema: uses feature definitions
ml/dataset.py - Model Training → Dataset Preparation: prepares data
ml/train_model.py - Model Training → Model Utilities: uses model utilities
ml/train_model.py - Model Evaluation → Dataset Preparation: evaluates splits
ml/evaluate_model.py - Model Evaluation → Model Utilities: uses metrics
ml/evaluate_model.py - Error Analysis → Dataset Preparation: analyzes data
ml/error_analysis.py - Error Analysis → Model Utilities: uses model utilities
ml/error_analysis.py - History Dashboard → API Routes: loads scan history
README.md - API Routes → Prediction Service: serves model details
backend/routes/model.py