# Code of Conduct

This project adopts the Contributor Covenant v2.1.
https://www.contributor-covenant.org/version/2/1/code_of_conduct/

## Additions specific to this project

Because this project handles reports of harmful AI behaviour, three rules go beyond the
Covenant:

1. **Never post unredacted content.** Not someone's personal information, credentials, or
   private data — including inside a screenshot of a model output.
2. **Working exploits and malware stay out.** Describe the pattern, file it through the
   report intake, do not publish a payload.
3. **Attack claims, never people.** Disagreement about whether a tag fired correctly is the
   entire point of the project. Contempt for the person holding the position is not.

## Enforcement

Public GitHub discussions and issues are moderated by repository maintainers.
Keep security and personal reports restricted. Appeals should receive review by
a maintainer not involved in the original action. A staffed appeal program and
quarterly statistics are planned, not operating commitments of this preview.

The owner created conduct@aitrustid.com; its delivery has not been independently
tested. Do not put private conduct reports in a public issue.