You file.
One problem, written the way you'd say it to a sharp colleague. Attach up to three links as evidence. Then close the laptop.
problem: "Should we enter the Gulf eSIM market, and how?"
deadline: 07:00 · budget armed
Scorpione takes one problem at nightfall, sends a pack of agents through the dark, and leaves a brief at your door by 07:00. You are not part of the hunt. You read what it killed.
File tonight's huntThis is one night on Scorpione — a real shape of a hunt, hour by hour. No chat. No prompts to babysit. The night does the work.
One problem, written the way you'd say it to a sharp colleague. Attach up to three links as evidence. Then close the laptop.
It breaks your problem into vectors of attack — the questions that, answered, would settle the matter.
Researchers run each vector in parallel, deep into sources. Every finding is written down with where it was caught — nothing enters the case without a trail.
One agent exists only to kill weak findings. It attacks the case until it runs out of objections — whatever dies here goes to the kill log, not to you.
What survived the night becomes one strike — a position you can act on — backed by the case, with every kill accounted for.
Three parts, always: the strike, the case, the kill log. If the night ran out of budget or trail, the brief says so — Scorpione reports an honest partial over a confident bluff.
Enter through travel eSIM in the UAE first — regulation is lightest and distribution is buyable. Do not lead with KSA: mandatory pre-registration makes the funnel die at install.
5 findings died in the night:"incumbents are slow to react" — single-source opinion"CAGR 41%" — contradicted by two filings
…and 3 more, each with cause of death.
Scorpione is not a chatbot and not a marketplace. Each hunt is a curated discipline, released when it survives our own nights.
File one problem before midnight. Read the brief with your coffee.
File tonight's huntOne email per release. No feeding schedule, no drip.