Sources: Ruben Hassid's LinkedIn post (2026-11-13) and his newsletter issue Prompt 5.2. (How to AI, 2026-11-13). Claims about caching and model behavior are attributed in the newsletter to Anthropic's docs and help center.
The advice
Use the newest, most expensive model (Claude Fable 5.2) for one big job: planning and architecting the whole "how". Then move execution to the cheaper everyday model (Claude Opus 5.5) through a written brief in a new chat, not by switching models mid-conversation. The newsletter adds that Fable's inclusion in standard Team seats is time-limited, after which it bills against usage credits. That makes the cost split practical, not theoretical.
Why not just switch models in the chat
Two reasons, both sound:
Caches are model-scoped. A long chat stays cheap because its prefix is cached. Switching models means the new model reads the whole history cold, at full input price. ("Rereads from zero" is the newsletter's phrase. Mechanically, it is a cache miss on the whole prefix.)
Reasoning doesn't transfer. The second model sees the visible transcript, not the first model's thinking. So the plan's rationale is lost unless it is written down.
The fix for both is the same: have the planning model write an explicit, self-contained brief, and start fresh.
The handoff brief (structure)
The newsletter's handoff prompt asks the planning model to write for a reader who saw none of the chat, copying exact names, numbers, dates, links and file names, keeping only final versions and adding nothing new. Its sections, paraphrased:
| Part | Purpose |
|---|---|
| Goal + "done when" | One sentence, then the checks that define success |
| Next step | The first action the executor can take without the human |
| Decided | Each final decision with a short reason, plus rejected options and why |
| Facts | Everything the work depends on |
| Done here | What is finished, so it isn't redone |
| Ask me first | What needs human approval, and why |
| Open | Only what is still undecided; no new questions |
| Checklist | Every remaining step as tick boxes |
| Executor rules | Read every relevant file and connected app first; work the checklist; keep going instead of offering to continue; stop only for approvals; don't reopen decisions but flag contradictions; end with a short recap |
The executor is then started in a new chat with the brief attached and a one-line instruction to take it from there.
Why it works, in engineering terms
This is context compaction done by hand, with a schema. The brief is a lossy but deliberate summary in which:
decisions carry their reasons and their rejected alternatives, so the executor doesn't relitigate them;
completion criteria are explicit, so "done" is checkable;
approval gates are named up front, so the executor knows when to stop;
state is split into done, open and remaining, which is what any resumable workflow needs.
The same structure works for handing off between agents, between sessions, or from an agent to a human.
Caveats
Lossy by design. Anything the planner leaves out of the brief is gone. Nuance that lived in the reasoning, or in rejected paths not listed, won't reach the executor. Review the brief before handing off.
The cheaper model may still need the expensive one. If execution hits a design question the plan didn't anticipate, the executor will improvise. The "Ask me first" and "don't reopen Decided" rules contain this, but don't prevent it.
"Keep going, don't offer to continue" cuts both ways. It reduces stalled sessions, but it raises the cost of a wrong turn. Pair it with explicit approval gates for anything destructive, as the brief does.
Benchmark framing in the newsletter is loose. It describes a 41.3% Remote Labor Index score as Claude doing better than paid humans 41% of the time. That benchmark measures the share of real freelance projects completed at an acceptable standard, not head-to-head wins over humans.
Pricing and plan details change. The specific inclusion window and limits are time-bound. The planner/executor split stays useful after they change.
Checklist for using the pattern
Does the brief state "done when" in checkable terms?
Are rejected options listed, with reasons?
Are approval gates named, especially for deletes, sends and spend?
Would a newcomer understand every name, path and number without the original chat?
Is the brief short enough that the executor reads all of it?
Bottom line
Practical, cost-aware advice with a reusable artifact: a structured handoff brief. The underlying reasons (model-scoped caches, reasoning that doesn't transfer) are correct. The brief template is worth keeping even if you never split work across models.