HAMMAD YOUSUF

FREE RESOURCE · FROM THE "BETTER" REEL

The make-it-better loop: Sonnet 5.5 vs Opus 5.5

By · 7 November 2026

One prompt, sent every round to a fresh Claude Code session. 10 rounds on Sonnet 5.5, 10 on Opus 5.5; a screenshot and snapshot after every round.

The prompt

The make-it-better loop prompt (sent every round).

prompt · claude code
Open the website first and look at it, then innovate. Put yourself in the customer's shoes and think beyond the obvious. Make ONE bold change that lifts conversions or the look and feel, not a minor tweak. You can use any tool you have. You must ship a change this round.

How we ran it

  • Round 0: a deliberately ugly one-page coffee site.
  • Each round: a fresh Claude Code session, same prompt, only the site folder.
  • 10 rounds on Sonnet 5.5, 10 on Opus 5.5; a screenshot and snapshot after every round.

What happened

Sonnet 5.5: 28 to 231 lines, careful ordering upgrades.

Opus 5.5: 28 to 1,089 lines, new features every round, invented customer reviews and a 4.9 rating.

Web searches: 0 for both.

Run it safely

Work on a copy, commit after every round, and cap spend (claude -p has --max-turns and --max-budget-usd). Claude Code needs a paid Claude plan or an API key.

Sources checked

  • Claude Code needs a Pro, Max, Ultra, Team, Enterprise or Console account, or an API key. Source ↗
  • Claude Code is not open source (all rights reserved). Source ↗
  • Sonnet 5.5 and Opus 5.5 are current Claude models usable in Claude Code. Source ↗
  • claude -p headless mode, --max-turns, --max-budget-usd. Source ↗

Our run: 10 rounds per model, line counts, zero web searches (experiment/summary.json, rounds.jsonl).

About the video: the reel is presented by an AI avatar of Hammad Yousuf.

Keep going