Exempt faithful preserve-by-default splits from the pre-merge gate#165
Merged
Conversation
Narrows the locked 'full regression re-run before merge' rule (RUBRIC_AUDIT.md / REGRESSION_GATE.md). A faithful preserve-by-default split — structural c2-split that only redistributes a locked old-c2=T into c2=T+c3=T, with zero realignments — is exempt: it changes no locked value's meaning, and since we don't act on grader output, a gate run can't change any expected value (it would only produce an informational Sonnet-calibration snapshot). Local validators + the faithful decomposition suffice. Gate stays REQUIRED for any change that realigns expected values or any new authoring. Recorded to avoid silent process drift (the kind that bit PR #146).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Records the decision to skip pre-merge gate runs for faithful preserve-by-default splits, so the revision is deliberate and documented rather than silent drift.
What changed
docs/REGRESSION_GATE.md— new section "When the pre-merge gate is required vs. optional":old-c2=T→c2=T+c3=T, zero realignments) — it can't change any expected value, so the gate would only produce a calibration snapshot. Local validators + the faithful decomposition suffice.docs/RUBRIC_AUDIT.md— one-line note on the locked "regression backfill" bullet pointing to the narrowed rule, so the two docs don't contradict.Why
The original "always gate before merge" rule predated the grader/gold-standard principle (PR #161). Its purpose was to catch grader miscalibration we'd act on; under preserve-by-default we don't act on grader output, so that purpose doesn't apply to these mechanical splits. Recording it avoids the silent-drift failure mode that produced the un-gated PR #146.
Test plan
After this lands: Units 7–9 proceed as faithful preserve-by-default splits, no gate runs.
Generated by Claude Code