Please report vulnerabilities privately through GitHub's "Report a vulnerability" form on this repository. Do not open public issues for security problems. Expect an acknowledgement within 5 business days.

• Model artifacts are serialised with joblib (pickle). Loading an artifact executes code, so only load artifacts you trained and stored yourself. Artifacts carry a SHA-256 manifest and can be HMAC-signed (AI_IDS_MODEL_HMAC_KEY); keep that key out of the repository.
• Never commit packet captures, datasets, API keys, or trained models.