Document GitHub moderation policies

github-management/github-moderation.md

@@ -0,0 +1,78 @@
+# GitHub Moderation
+
+Contributions to Kubernetes GitHub repositories must meet legal, technical, and behavioral standards. Submissions that fail to meet these expectations, whether due to copyright violations, low-quality automated generation, spam-like behavior, or bad-faith engagement may be moderated, restricted, or escalated according to established project processes.
+
+GitHub moderation is handled at the repository level by maintainers, with escalation to **[@kubernetes/owners](https://github.com/kubernetes/org/blob/main/OWNERS)** and the [Kubernetes Code of Conduct Committee](./committee-code-of-conduct) when necessary.
+
+All actions must follow the [Kubernetes Code of Conduct](./committee-code-of-conduct) and the escalation procedures defined in the [moderation documentation](/communication/moderation.md).
+
+## Comment Moderation
+
+Maintainers may edit or remove comments that:
+
+- Violate the [CNCF Code of Conduct](https://github.com/cncf/foundation/blob/main/code-of-conduct.md).
+- Contain harassment or personal attacks
+- Include sensitive personal information
+- Are spam or malicious
+
+## Legal and Copyright Compliance
+
+All contributions must:
+
+- Be covered by a signed [Contributor License Agreement (CLA)](https://git.k8s.io/community/CLA.md).
+- Comply with copyright requirements.
+- Not introduce code copied from third-party sources in violation of license terms.
+
+This applies equally to:
+
+- Human-authored content
+- Copy-pasted content
+- AI-generated content
+
+Use of generative tools does not exempt contributors from CLA or copyright obligations. Contributors are responsible for ensuring they have the right to submit all material.
+
+See [Linux Foundation guidance](https://www.linuxfoundation.org/legal/generative-ai) on generative AI and licensing expectations.
+
+## Quality and Engineering Standards
+
+Each pull request requires reviewer time and project resources.
+Submissions must:
+
+- Demonstrate understanding of the change being proposed.
+- Be technically sound.
+- Meet project quality standards.
+- Be responsive to review feedback.
+
+Contributions that appear automated, mechanically generated, or submitted without understanding may be closed if:
+
+- The submitter cannot meaningfully respond to review.
+- The change introduces avoidable defects.
+- The review burden is disproportionate to the value of the change.
+
+This standard applies regardless of whether content was human-written or AI-generated. Contributors must understand and be able to explain and modify the changes they submit, regardless of how they were authored.
+
+## Spam, Automation, and Bad-Faith Contributions
+
+Kubernetes has historically received automated and low-quality submissions. Repeated spammy or obviously automated behavior may result in:
+
+- PR closure
+- Issue locking
+- Participation restriction
+- Escalation to organization-level moderation
+- Account banning where appropriate
+
+Excessive low-quality automated submissions are treated as disruptive behavior.
+
+Moderation decisions will consider:
+
+- Volume of submissions
+- Responsiveness to feedback
+- Pattern of behavior across repositories
+
+## Trivial or Fragmented Edits
+
+Contributors are encouraged to batch related fixes. Trivial single-line edits that create unnecessary review overhead may be closed with a request to consolidate improvements into a single submission.
+
+Automation reduces cost, but review remains an engineering effort. Contributors should respect reviewer time.
+
+See [guidance on trivial edits](https://www.kubernetes.dev/docs/guide/pull-requests/#trivial-edits) in the pull request guidelines.