One-Shot Shop Challenge

The first build on Fable 5, the Mythos-class model Anthropic released publicly on the same day this run started. Same one-prompt brief and the same ~50-page spec as the Opus 4.8 run (#13) right before it, with one deliberate change: where #13 was ordered into strict team mode, this run left the orchestration up to the model. The brief said to build the whole shop in one go and test it, but never how to structure the work.

The first Opus 4.8 build on the page, and a direct descendant of the baseline build #1. Same idea - one team-mode agent, one one-shot prompt, the same ~50-page spec - but two model generations on (Opus 4.6 to 4.8) and dialled all the way up: xHigh reasoning, thinking on, 1M context, and a hard instruction to “implement the entire shop in one go without stopping, use team mode, and test everything via Pest.” The question was what those upgrades actually buy over the original team run.

Same model and CLI family as build #11, but driven by Codex 0.128’s new persistent goal mode: a single multi-page brief is wrapped in an <untrusted_objective> tag and the agent loops on its own through PLAN → IMPLEMENT → VERIFY → INDEPENDENT QA → FIX → COMMIT until it calls update_goal("complete"). Reasoning was set to xHigh.

First run on GPT-5.5 in Codex CLI. The user prompt this time explicitly demanded sub-agent role-play and Pest + Playwright testing in one go.

Rerun of build #9 with two changes: the prompt was hardened (“have to” replaced “must” on every rule so the instruction to use team-mode could not be rationalised away) and reasoning was raised to xHigh, which the model reportedly handles better than High. All third-party plugins, including the design plugin, were unplugged first so the output reflects Opus on its own.

Same prompt as build #1, just swapped to Opus 4.7. The run finished in 32m 27s for $22.63, by far the fastest on this page (build #1 took a full hour).

Same specification, same technology, and this time the same original prompt as build #1. The idea was to check whether the increased 1M token context limit plus the 42 Claude Code releases since the first run (v2.1.39 to v2.1.81) produce a better result on their own, without any prompt engineering.

Same specification, same technology, but this time with an advanced prompt combining several techniques: a thorough controller supporting the team lead, BDD and TDD, code reviews, and a dedicated QA teammate that actively tries to break the system. The prompt was designed to leverage the new 1M token context window.

Claude Code took a phased approach. The team lead read the full specification, broke it into 12 implementation phases with explicit dependencies, and then spawned specialized agents for each area: migrations, models, Livewire components, admin panel, seeders, and so on.

Same specification, same prompt structure, but this time Claude Code ran with sub-agents instead of team mode. No specialized agent instructions were prepared - the prompt simply told it to use sub-agents. Claude spawned 20 sub-agents total, 12 of which actively contributed code.

Codex launched explorer agents to analyze the specification first, synthesized their findings into a phased roadmap, then delegated implementation to worker agents. The process took about 1 hour and 44 minutes with 16 sub-agents total.

Same specification, same technology, but this time with a tuned prompt and strict quality constraints. The prompt included mandatory PHPStan compliance at max level, Deptrac architectural boundary checks, Pest test coverage, QA self-verification against every acceptance criterion, and a fresh agent review cycle where a new agent instance re-evaluated the entire codebase.

Same specification, same technology, but this time Codex received custom instructions with two additional quality tools: PHPStan (static analysis at max level) and Deptrac (architectural boundary checks). The idea was to see if giving Codex explicit quality constraints would produce measurably better code.