prove-it

Claude generates code by pattern-matching on training data. Something can look syntactically perfect, follow best practices, and still be wrong. The model optimizes for "looks right" not "works right." Verification is a separate cognitive step that must be explicitly triggered. This skill closes the loop between implementation and proof.

1. Generation vs Execution I generate code but don't run it. I predict what it would do based on patterns. My confidence comes from "this looks like working code I've seen" not from "I executed this and observed the result."

2. Training Signal Mismatch My training optimizes for plausible next-token prediction, not outcome verification. Saying "Done!" feels natural. Verifying feels like extra work.

قبل الإعلان عن اكتمال أي مهمة، تحقق فعليًا من النتيجة. قم بتشغيل الكود. اختبر الإصلاح. تحقق من الإخراج. يعمل تدريب كلود على تحسين المظهر المعقول الإخراج، لم يتم التحقق من الإخراج الصحيح. تفرض هذه المهارة خطوة التحقق هذا لا يأتي بشكل طبيعي. ولا انتصار ينتصر دون برهان. المصدر: elliotjlt/claude-skill-potions.