Open-source subset of Shieldyn focused on rule-based phishing detection, URL parsing utilities, and an evaluation harness.
- Rule-based detection heuristics
- URL extraction and shortener detection
- Optional Safe Browsing checks (no keys included)
- Evaluation harness with synthetic samples
- OAuth, billing, multi-tenant infrastructure, deployments
- AI model weights or inference
- Customer data, internal endpoints, or secrets
python -m venv .venv
source .venv/bin/activate
pip install -e .
python -m shieldyn_oss.eval data/samples.jsonl
from shieldyn_oss.rules import RuleBasedDetector
sample = "Urgent: verify your account"
result = RuleBasedDetector().analyze(sample, subject="Action required")
print(result.score, result.indicators)The Safe Browsing helper is async and requires your own API key.
import asyncio
from shieldyn_oss.safebrowsing import check_urls_safe_browsing
urls = ["https://example.com"]
flagged = asyncio.run(check_urls_safe_browsing(urls, api_key="YOUR_KEY"))
print(flagged)pip install -e ".[dev]"
pytest
If you discover a security issue, see SECURITY.md.
Apache-2.0. See LICENSE.