feat(skills): 'one symbol, one meaning' guardrail (df-conflation trap)#112
Merged
Conversation
…ath-tutorial Encodes the systemic lesson from the Missing-Effect adversarial review: a 4th closure-audit rule against attaching one definition of a symbol to a reference that needs a different form. Headline trap — a degrees-of-freedom defined as the scalar Rubin (m-1)(1+1/r)^2 then tied to a Meng-Rubin/LRR pooled-LRT reference (numerically ~62 vs ~48.5). Also: never equate a quantity with a function of itself (ARIV is r, not 1/(1+r)). Audit algorithm step (e) extended to check per-context disambiguation. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Encodes the one systemic lesson from the Missing-Effect adversarial review into the
scaffolded-math-tutorialskill so future/tutorruns don't re-introduce it.The trap: every
/tutor-generated tutorial conflated two different degrees-of-freedom — the scalar Rubin(m−1)(1+1/r)²(≈62 for the headline cell) vs the Meng–Rubin/LRR pooled-LRT df (≈48.5) — by definingνonce and attaching it to a reference that needs the other form. The hand-built proof tutorial got it right; the generated ones regressed.The fix: a 4th Terminology-Closure-Audit rule — one symbol, one meaning; flag context-dependent forms — with the df conflation as the worked example, plus a note never to equate a quantity with a function of itself (ARIV is
r;1/(1+r)is a function of it). Audit-algorithm step (e) extended to check per-context disambiguation.Docs-only change to one SKILL.md (still well under the 500-line budget).
🤖 Generated with Claude Code