How aditiya-saputra/brantas-judol works
Brantas-Judol discovers candidate domains from Certificate Transparency logs, enriches candidates with urlscan evidence, and experimentally scores them with Gemini. Its output is explicitly candidate signals, not a human-reviewed blocklist. The CT parsing and cursor details are in one script; the graph separates their domain stages without implying separate modules. Probe scan submission is an optional path. Some relationships and capabilities are documented in the README; unsampled implementation details are not inferred.
Main entry points: classify.py Candidate classifier · probe.py Evidence probe · grab.py CT grabber · urlscan-evidence.json Evidence ledger
Components 17
CT discovery
- CT grabber
grab.py - Log fetch
grab.py - TLS and X.509 parsing
grab.py - Domain normalization
grab.py - Keyword filter
grab.py
Evidence probing
- Evidence probe
probe.py - Scan submission
probe.py
LLM classification
- Candidate classifier
classify.py - Score buckets
Persistent state
- Candidate domains
- CT cursors
ct.db - Evidence ledger
urlscan-evidence.json
Other components
- Operator
- CT registries
- urlscan.io
- Gemini API
- Google Search grounding
Connections 21
- Operator → CT grabber: runs scan
README.md - CT registries → CT grabber: provides log list
README.md - CT grabber → Log fetch: dispatches fetch
grab.py - Log fetch → CT registries: fetches entries
README.md - Log fetch → TLS and X.509 parsing: parses entries
README.md - TLS and X.509 parsing → Domain normalization: extracts domains
README.md - Domain normalization → Keyword filter: filters domains
README.md - Keyword filter → Candidate domains: appends candidates
README.md - CT grabber → CT cursors: reads and updates
README.md - Operator → Evidence probe: runs probe
README.md - Evidence probe → Candidate domains: reads candidates
probe.py - Evidence probe → urlscan.io: searches scans
probe.py - Evidence probe → Scan submission: optionally submits
probe.py - Scan submission → urlscan.io: submits scan
probe.py - Evidence probe → Evidence ledger: writes evidence
README.md - Operator → Candidate classifier: runs classifier
README.md - Candidate classifier → Candidate domains: loads candidates
classify.py - Candidate classifier → Evidence ledger: loads evidence
README.md - Candidate classifier → Gemini API: requests scores
classify.py - Candidate classifier → Google Search grounding: optionally grounds
README.md - Candidate classifier → Score buckets: writes score buckets
README.md