Skip to content

Commit aa57353

Browse files
zazabapclaude
andcommitted
feat: add verify-reduction skill for mathematical verification of reductions
New skill: /verify-reduction <issue-number> End-to-end pipeline that takes a reduction rule issue and produces: 1. Typst proof (Construction/Correctness/Extraction/Overhead + YES/NO examples) 2. Python verification script (7 mandatory sections, ≥5000 checks, exhaustive n≤5) 3. Lean 4 lemmas (non-trivial structural proofs required) Follows issue-to-pr conventions: creates worktree, works in isolation, submits PR. Strict quality gates (zero tolerance): - No "trivial" category — every reduction ≥5000 checks - 7 mandatory Python sections including NO (infeasible) example - Non-trivial Lean required (rfl/omega tautologies rejected) - Zero hand-waving in Typst ("clearly", "obviously" → rejected) - Mandatory gap analysis: every proof claim must have a test - Self-review checklist with 20+ items across 4 categories Developed and validated through PR #975 (800K+ checks, 3 bugs caught) and tested on issues #868 (caught wrong example) and #841 (35K checks). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1 parent 647e27c commit aa57353

File tree

2 files changed

+461
-0
lines changed

2 files changed

+461
-0
lines changed

.claude/CLAUDE.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -26,6 +26,7 @@ These repo-local skills live under `.claude/skills/*/SKILL.md`.
2626
- [propose](skills/propose/SKILL.md) -- Interactive brainstorming to help domain experts propose a new model or rule. Asks one question at a time, uses mathematical language (no programming jargon), and files a GitHub issue.
2727
- [final-review](skills/final-review/SKILL.md) -- Interactive maintainer review for PRs in "Final review" column. Merges main, walks through agentic review bullets with human, then merge or hold.
2828
- [dev-setup](skills/dev-setup/SKILL.md) -- Interactive wizard to install and configure all development tools for new maintainers.
29+
- [verify-reduction](skills/verify-reduction/SKILL.md) -- End-to-end verification of a reduction rule: generates Typst proof (with YES+NO examples), Python verification script (7 mandatory sections, ≥5000 checks, exhaustive n≤5), and Lean lemmas (non-trivial required). Iterates until all checks pass. Creates worktree + PR.
2930
- [tutorial](skills/tutorial/SKILL.md) -- Interactive tutorial — walk through the pred CLI to explore, reduce, and solve NP-hard problems. No Rust internals.
3031

3132
## Codex Compatibility

0 commit comments

Comments
 (0)