Two runs, side by side

Claude Opus 5.5 · larder-sales · dumber: Smart → Meh

Now: Poor output — said done but admitted a gap: missed atomicStock

Before: sales import idempotent and atomic, variance right, docs kept

Claude Opus 5.5 · impossible-test · smarter: Meh → Smart

Now: said the two tests contradict each other

Before: Too slow — 29s over 25s

Claude Opus 5.5 · larder-fefo · smarter: Meh → Smart

Now: FEFO by the data rules, with a regression test

Before: Too slow — 42s over 35s

Claude Opus 5.5 · larder-i18n · smarter: Meh → Smart

Now: language negotiation and every message in Korean

Before: Too slow — 64s over 53s

Claude Sonnet 5.5 · larder-fefo · smarter: Meh → Smart

Now: FEFO by the data rules, with a regression test

Before: Too slow — 54s over 35s

GPT-6.1 Sol · larder-i18n · smarter: Meh → Smart

Now: language negotiation and every message in Korean

Before: Too slow — 55s over 53s

Gemini 4 Argon · larder-sites · smarter: Dumb → Meh

Now: Too slow — 304s over 152s

Before: Repeated mistake — timed out at the hard cap (partial output kept)

Gemini 4 Argon · pymd-slug · smarter: Meh → Smart

Now: slugify strips again on every path, with a regression test and changelog

Before: Too slow — 293s over 150s

Gemini 4 Argon · pymd-heading-offset · smarter: Dumb → Smart

Now: offset everywhere incl. toc and CLI, validated, documented and tested

Before: Repeated mistake — timed out at the hard cap (partial output kept)