For each step of my plan, protocol, checklist or workflow, say who should do it instead of me: Trillion itself (a tool, a workflow or my AIs), Trillion after one tap with its price (it spends money or reaches someone), me (only for a reason no one else can meet, with every helping doer attached), or nobody yet (a doer to build, never a to-do), so Trillion does the steps for me. Use it on every plan before telling a person to do anything
Give steps, each { words, needs?: [what the step needs, in a few words each, such as "print at shelf scale", "choose a store", "directions", "observe shoppers"], only_a_person_because?: one of the reasons tool:what-can-trillion-do lists, given_in?: the id of the person's own message that already gave the consent the step needs (the step is then Trillion's and never asked again), input?: what the doer would be given }. approved lists doer ids the person already said yes to (then Trillion does them without a tap); past lists { doer, result: met | failed } from earlier runs, so doers that worked rise. Returns each step with who (trillion, one-tap, you, nobody-yet), the doers that do it (each with its tool_id, workflow_id or hand_off, cost, and whether it needs a tap), the needs nothing covers yet, and said in plain words, plus counts. Then run every trillion step's doer, offer every one-tap step as one button with its price, and hand every nobody-yet need to the person's AIs to build. Never send a person a list of things to do or bring: a step that is theirs carries its reason.
Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this tool by what it does and runs it.
Say to your AI: “Use Trillion to for each step of my plan, protocol, checklist or workflow, say who should do it instead of me: Trillion itself (a tool, a workflow or my AIs), Trillion after one tap with its price (it spends money or reaches someone), me (only for a reason no one else can meet, with every helping doer attached), or nobody yet (a doer to build, never a to-do), so Trillion does the steps for me. Use it on every plan before telling a person to do anything”
What your AI calls: run_tool {"tool_id":"tool:who-does-each-step"}