Sunday Coffee & Code: When the New Models Team Up
This week’s experiment: my website needed a rework, and I decided to make it a two-model job.
The setup:
- ChatGPT Astra (running on high) did the analysis — a full review of the site and a detailed markdown file of recommended changes
- Claude Code running Fable 5.1 (also running on high) took that md file and did the work
The result? A reworked website, including a revolving logo with three rotating versions that I didn’t even know I wanted until I saw it.
The interesting bit for me, as this wasn’t really about my website. It was about seeing what the new generation of models can do together - one as the analyst, one as the builder, with a markdown spec as the handoff artifact between them.
That handoff pattern is essentially spec-driven development with a mixed-vendor toolchain: the analysis model doesn’t need to know how the implementation model works, and vice versa. The spec is the contract.
The economics: Fable 5.1 came in at $44.67 (and took over an hour - with a bit of back and fore) for the whole build. For a full website rework plus an animated logo? More than acceptable.
My website has always done double duty as part of my development and experimentation platform - every rework is also a test bench. This one confirmed something I’ve been thinking on, the interesting capability gains aren’t just in individual models. They’re in how you orchestrate them.
𝗕𝗮𝗰𝗸 𝘁𝗼 𝘁𝗵𝗲 𝗰𝗼𝗳𝗳𝗲𝗲 - 𝗿𝗲𝗺𝗲𝗺𝗯𝗲𝗿: 𝗧𝗵𝗶𝗻𝗸. 𝗠𝗼𝗱𝗲𝗹. 𝗔𝗰𝘁. ☕