An AI strategy consulting team,tailored to your business.
Set up your business and the problem you’re trying to solve, and we’ll source a specialist team and build a tailored playbook around it. You run the prompts yourself, in your own tools and on your own timetable, for a fraction of what a consulting engagement would cost.
How it works
Five steps. You can read all 124 tool pages without an account. With one, everything below is free to set up, and three tool prompts a month are free to run.
1 — Tell it about your business
Name the business and say what it does. Two answers is enough to start, or copy a nine-question interview to run in your own Claude and paste back. Free.
2 — Say what has to be settled
One sentence on the decision this engagement has to reach. An eight-question version is there to run the same way if the decision isn’t clear yet. Free.
3 — Brief the engagement partner
Whitmore, until you rename him, reads the business and the engagement, works out which specialisms the work needs, and builds specialists with depth in your sector. Sourcing them is free. Standing them up is one paste into your own Claude Code, where they become real subagents with a model pinned to each, and that part is paid.
4 — Build the playbook
Diagnose, Explore, Choose, Plan, Review, with the tools inside each stage taken from your brief. Assembling it yourself from the bank is free. Having the workbench build the sequence is paid.
5 — Run it side by side
You work in your own Claude Code with the workbench open beside it, calling tools through your partner or pasting one self-contained prompt at a time. Paid both ways, and your work stays in your project.
Example team
Whitmore, your engagement partner, plus seven of the specialists he could source — the same cards the real team page uses.
The bank
124 tools across 12 categories, each with what it is, what it is for, what it is not for, and five steps for running it as a workshop.
Built for Claude Code in the Claude desktop app, because it is the only place a consulting team can be stood up as real subagents with a model pinned per consultant. That is the reason for narrowing, not a limitation of it.