TensorFold uses a simple standard for project participation: be direct, be technical, and be respectful.
- Assume good intent while still checking claims carefully.
- Critique code, measurements, and designs rather than people.
- Make performance claims reproducible.
- Keep discussions focused on the runtime, benchmarks, documentation, and user experience.
- Harassment, threats, or personal attacks.
- Publishing private information such as local paths, credentials, email addresses, or machine-specific details without consent.
- Repeatedly presenting unverified benchmark claims as facts.
- Deliberately submitting code or artifacts that hide unsafe behavior.
Maintainers may edit, hide, or remove comments, issues, pull requests, or contributions that violate this code. Serious or repeated violations may lead to blocking from the project.