Coming Soon · Free
Before You Point AI At It — pressure-test the move you've already decided to make.
A guided pre-flight read for the operator who has already decided to point AI at something — rebuild the CRM in Claude Code, replace the quoting tool, automate an approval flow, let the AI redo onboarding. Friday does not talk you out of it. You bring one real move; Friday runs it as your own pre-mortem, walks the five things your move quietly touches that the AI cannot see, and the prep gap falls out of your own blast radius — scored against a fixed per-context bar, never lectured. The cost lands in your own numbers, and the verdict is honest: go now, do these first, or stop. You leave with a re-runnable Prep Gap Read. If the radius is genuinely small, Friday waves you through — zero workshops, build it now.
The arc
A short calibration read — how you run the place, and the one AI move you're closest to actually doing. The named move is the anchor the whole session hangs off.
State the move concretely — what the AI does, what gets replaced, what 'done' looks like — and Friday reflects it back so you own the framing. The cited 'context engineering' frame lands here, once, on your move.
The spine. An operator-owned pre-mortem: imagine the move already failed and trace why, walking the five threads — downstream, upstream, data/integrations, dependent people, encoded rules — one at a time to a named specific. Closes on your priced worst single run.
Also from Framework Friday
What you'll walk away with
Score your workflows against five operator-tier criteria, map your top priority end-to-end, and produce a spec with governance that a builder could pick up tomorrow. Yours to keep, refine, and put to work — whichever way you choose to run it.
A derivative verdict pass: the gap falls out of your blast radius. Friday sorts the threads into the four hidden contexts and scores each against a fixed LOADED / PARTIAL / MISSING bar, each tied to a real attributed failure. Emits a stamped, re-runnable ledger header.
Completes the four-number engine the worst-run pre-mortem seeded in phase 3 — in your own figures, never Friday's. Worst single run (sharpened), rebuild hours, rework hours + who eats it, trust hit. Then the honest prep-cost vs skip-cost net-out.
For each context verdicted partial or missing against a named failure, the specific live workshop that closes it, in run-order, each with a one-line reason tied to your move. LOADED contexts route to nothing. All-loaded → empty plan → go now.
The one-sentence honest verdict — go now / prep-first / stop — the four binary completion screens run out loud, the Prep Gap Read exported with its stamped ledger header, the roadmap-completion signal fired, and the FridayOS bridge as continuity.