🏷️ Discussion Type

Product Feedback

Body

Summary

Add a toggle in the pull request section of branch rulesets to allow authors to explicitly approve their own commits, enabling regulatory compliance for mixed human-AI development workflows without creating unnecessary friction.

Problem Statement

GitHub's branch protection rules prevent authors from approving their own PRs. This design works well for human-only development but creates friction in hybrid human-AI workflows:

• Regulatory Requirement: We must have 2 human reviewers involved in all production changes
• Current Practice: Small teams (2 people) enforce 1 required review + rely on implicit author approval (the author doesn't formally approve, but authorship implies approval)
• The AI Problem: AI agents cannot count as human reviewers. An AI-authored change reviewed by 1 human = only 1 human reviewer (fails requirement)
• Current Workaround: Increase required reviews to 2 for everyone, creating unnecessary friction for human developers who previously satisfied the second reviewer via implicit author approval

Proposed Solution

Add a checkbox in the Pull request section of branch rulesets:

☐ Allow authors to approve their own pull requests

When enabled:

• Authors can explicitly approve their own PRs (converting implicit approval to explicit)
• Author approval counts toward minimum required approvals
• Self-approvals are visible in PR timeline for auditability

When disabled (current behavior):

• Authors cannot formally approve their own PRs (implicit approval only)
• Existing rulesets unchanged

Why This Matters

With this toggle enabled, the same rule works for both human and AI commits:

• Human: 1 explicit reviewer + 1 explicit author approval = 2 human approvals ✓
• AI agent: 2 explicit human reviewers (AI implicit approval doesn't count for compliance) ✓

This eliminates the need to:

• Increase required reviews to 2 for all human commits (unnecessary friction during gradual AI adoption)
• Maintain separate rulesets for AI vs. human commits (complex and hard to audit)

Benefits

1. Regulatory compliance — Explicit audit trail of all 2 required approvals
2. Reduced friction — Same rules apply to both human and AI workflows
3. Simple to maintain — Single consistent ruleset, no special exceptions
4. Security preserved — Still requires 2 approvers; only changes whether author approval is implicit or explicit