What problem does it solve? When a chapter's physics code or content changes, you need to re-validate correctness without rebuilding the whole chapter. This Skill re-runs only the verification phase—tests on the V1-V7 evidence ladder, benchmark references, and report generation—so regressions are caught quickly. ## Core Features & Use Cases - Targeted Re-Verification: Briefs the math-verifier on the chapter's evidence ladder tests proportional to curation tiers, then reports the verdict, per-label counts, and observed convergence orders. - Implementer Fix Loop: On FAIL, loops with the concept-implementer (max 3 iterations) applying physics-only fixes while keeping public signatures stable. - Post-Fix Consistency Checks: Re-runs explainer parity (tools/shot.py --quick) and notebook execution (tools/run_notebook.py) since physics changes may affect outputs, then updates progress.json and commits. - Use Case: After editing a chapter's solver, run /verify-chapter 7 to confirm all tests still pass, explainer screenshots still match, and the notebook still executes end-to-end. ## Quick Start Run /verify-chapter 7 to re-verify chapter 7's tests, fix any physics failures, and re-check explainer parity and notebook execution.
