How harshil-2008/phisheye works
PhishEye detects brand lookalikes in individual domains or new-domain feeds. The main workflow ingests a domain, screens its URL and registration age, then—when applicable—compares fetched page content and rendered appearance against brand snapshots. The scanner fuses available signals into a probability, verdict, and evidence; CLI/API entry points expose analysis and reporting. Training, brand snapshot refresh, and report generation are additional documented capabilities. Edges reflect the shown scanner orchestration, imports, or README; unsampled internals are represented only where documented.
Main entry points: scanner.py Scan orchestration · run.py CLI commands · similarity.py Page snapshots and similarity · whois.py RDAP age lookup
Components 17
Entry points
- CLI commands
run.py - REST API
phisheye/api.py
Screening
- Scan orchestration
phisheye/scanner.py - RDAP age lookup
phisheye/whois.py - URL features
phisheye/features/url_features.py - Screening classifier
phisheye/ml/model.py
Deep analysis
- Page snapshots and similarity
phisheye/content/similarity.py - Visual comparison
phisheye/visual/similarity.py - Score and verdict fusion
phisheye/fusion.py
Models and data
- WHOIS cache
data/whois_cache.json - Synthetic training data
phisheye/ml/dataset.py - Brand configuration
phisheye/config.py - Brand definitions
data/brands.json - Brand snapshots
data/brands_cache
Results
- JSON CSV HTML reports
phisheye/report.py - Generated reports
data/reports
Other components
- Analyst or feed
Connections 23
- Analyst or feed → CLI commands: runs commands
README.md - Analyst or feed → REST API: submits analysis
README.md - CLI commands → Scan orchestration: checks or scans
run.py - REST API → Scan orchestration: routes analysis
phisheye/api.py - Scan orchestration → URL features: extracts features
phisheye/scanner.py - Scan orchestration → Screening classifier: gets screening score
phisheye/scanner.py - Scan orchestration → RDAP age lookup: looks up age
phisheye/scanner.py - RDAP age lookup → WHOIS cache: reads and writes
phisheye/whois.py - RDAP age lookup → Analyst or feed: queries RDAP
- Screening classifier → Synthetic training data: trains from
phisheye/ml/model.py - Synthetic training data → URL features: extracts features
phisheye/ml/dataset.py - Scan orchestration → Page snapshots and similarity: compares snapshots
phisheye/scanner.py - Scan orchestration → Visual comparison: renders candidate
phisheye/scanner.py - Page snapshots and similarity → Visual comparison: hashes favicons
phisheye/content/similarity.py - Scan orchestration → Score and verdict fusion: fuses signals
phisheye/scanner.py - URL features → Brand configuration: loads brand tokens
phisheye/features/url_features.py - Page snapshots and similarity → Brand configuration: uses brand metadata
phisheye/content/similarity.py - Brand configuration → Brand definitions: loads definitions
phisheye/config.py - Scan orchestration → Brand snapshots: uses snapshots
README.md - CLI commands → Page snapshots and similarity: refreshes snapshots
run.py - CLI commands → Screening classifier: trains classifier
run.py - CLI commands → JSON CSV HTML reports: writes scan reports
run.py - JSON CSV HTML reports → Generated reports: writes outputs
phisheye/report.py