Intel / article

Moonlight & Mayhem (Raccoon Heist by Codex + GPT-5.6 Sol Ultra)

Source: simonwillison.net
Author: Simon Willison
Date: Published Aug 7, 2026

Context

Simon Willison reused the prompt from the earlier Raccoon Heist experiment with OpenAI's Codex and GPT-5.6 Sol Ultra. Willison judged the resulting museum-heist game better, one where rescuing other raccoons enables the player to steal a golden sardine. The account reports 52 minutes of work.

The initial version also gave each raccoon an enormous floating eyeball. Codex had reviewed screenshots without correcting it, and Willison needed two follow-up prompts to get a fix. The example shows an agent producing a substantial playable project while still needing a person's visual judgment to catch an obvious defect.

Terms in this piece · Glossary

AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
subagent — A separate agent spawned by another to do one scoped piece of work in its own context, returning only the result.

Key quotes

“I decided to pose the exact same prompt to Codex Desktop running GPT-5.6 Sol Ultra - the mode where Sol makes aggressive use of sub-agents - to see how it would do. It produced a much better game!”

“GPT-5.6 Sol has you in a museum, rescuing your two other raccoon crewmates in order to stack on top of each other and bust the golden sardine out of its case. Much more heisty!”

“Despite reviewing screenshots during development Codex failed to spot and correct this bug.”

Read the source simonwillison.net