The Russian IT Community Corpus project is committed to ensuring zero personally identifiable information (Zero-PII) in all public datasets and maintaining secure data processing pipelines.

If you discover any unmasked personal identifier (real full name, unredacted phone number, personal email, private API token, or cryptocurrency private key) in any published dataset artifact:

1. Do not create a public GitHub issue with the sensitive data.
2. Send an email with details (file name, row index / msg_id, and description of the leak) to the maintainers or open a PII Takedown Request.
3. SLA: Our team will acknowledge receipt within 24 hours and remove the flagged record from the dataset within 48 hours.

If you identify a security vulnerability in the pipeline code (e.g. arbitrary code execution, path traversal in export loaders, or dependency vulnerabilities):

1. Report the vulnerability via GitHub Security Advisories or maintainer contact.
2. Please provide a minimal reproducible proof-of-concept (PoC).
3. We will triage and release a security patch promptly.

All dataset releases undergo automated heuristic pattern checks, NER regression audits, and stratified probabilistic validation with Wilson score upper bounds before publication. The probabilistic audit report is generated in reports/probabilistic_pii_audit.json.