Please report suspected vulnerabilities privately through GitHub's security advisory form:
Include, when available:
- the affected repository, component, and version or commit;
- a description of the vulnerability and its potential impact;
- steps to reproduce or a minimal proof of concept;
- any known mitigations or workarounds; and
- a safe way to contact you for follow-up.
Do not include secrets, personal data, or data belonging to other users in a report. Use synthetic or redacted evidence whenever possible.
Please do not disclose an unremediated vulnerability in a public issue, discussion, pull request, or Discord channel. If the private reporting form is unavailable, contact an organization owner through GitHub and ask for a private reporting channel without sharing vulnerability details publicly.
We will make a reasonable effort to:
- acknowledge the report and establish a private communication channel;
- investigate its validity, severity, and scope;
- coordinate remediation and disclosure with the reporter; and
- publish an advisory when that would help affected users.
Response and remediation times depend on severity, complexity, and maintainer availability. Please allow a reasonable period for investigation and remediation before disclosure, and coordinate disclosure timing with us.
This policy applies to repositories maintained by the nrd-ai GitHub organization. Third-party services and dependencies should generally be reported to their respective maintainers unless the issue arises from how an nrd-ai project uses or configures them.
Reports made in good faith to improve security are welcome. This policy does not authorize privacy violations, service disruption, social engineering, destructive testing, or access to data beyond what is necessary to demonstrate the issue.