This repository publishes data only (no executable code). The most likely "security" concerns here are data-integrity or privacy issues — for example, a record that looks like it could identify a real person, a corrupted or malformed data file, or a licensing/attribution problem.
Do not open a public GitHub issue for a privacy concern. Instead:
- Email: send a detailed report to opensource@gruposantander.com
- GitHub Security Advisories: alternatively, use GitHub Security Advisories to report privately.
- Which dataset and which record(s) are affected (e.g.
id: SGCD-0123). - A description of the concern (privacy, integrity, license, attribution).
- Why you believe it is a problem and any supporting context.
| Stage | SLA |
|---|---|
| Acknowledgment of report | < 48 hours |
| Initial assessment | < 7 days |
| Remediation for privacy/integrity issues | < 30 days |
All datasets in this collection are derived from publicly available, already de-identified source datasets (e.g. the UCI Machine Learning Repository). The records do not contain Santander customer data, internal data, or newly collected personal data. The "stresses" applied are controlled, documented perturbations of the observed inputs — the underlying source record is never altered.
If you believe any published record could nonetheless identify a real individual, report it privately using the channels above and we will investigate and, if needed, withdraw the affected records.
This policy covers only the data and documentation in this repository. It does not cover Santander's internal infrastructure, other Santander products, or the original upstream source datasets (report those to their respective maintainers).
- Never commit secrets, API keys, tokens, or credentials.
- Never commit internal URLs, IP addresses, or corporate email addresses.
- Never add records derived from non-public or personally identifiable data.
We follow a coordinated disclosure process and ask that you give us reasonable time to remediate before any public disclosure.