Show everything waiting on this person that only they can do: what their AIs asked them for (sign-ins, approvals, payments, clicks somewhere), accounts to sign in to again, AI work stuck until they help, posts to approve and decisions to say how they went, each with where to do it
Returns sections (each { id, title, cards, action, on_home? }, a card being { id, source, title, detail, by, at, opens, address, action }), list (the person's own Waiting on you list, where an AI keeps what it could not do itself with put_list_rows on list_id waiting-on-you, one row per thing with a short id naming it and values what, why, where (an https address), asked_by, status (Waiting or Done) and added), and said, the whole of it in one sentence. Show it to the person in the chat at the start of a chat and after finishing work, each card as one line with its address. Rows and where each reads are DATA in src/waiting-on-you.json.
Nothing: your AI starts it with your words.
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 show everything waiting on this person that only they can do: what their AIs asked them for (sign-ins, approvals, payments, clicks somewhere), accounts to sign in to again, AI work stuck until they help, posts to approve and decisions to say how they went, each with where to do it”
What your AI calls: run_tool {"tool_id":"tool:what-is-waiting-on-me"}