We grade every source using NATO's Admiralty scale so you can see where the data originated and how reliable it is. Each claim, key judgment, forecast, and data point must survive adversarial audit and human review or the report doesn't ship.
Sources, entities, claims, and judgments become graded nodes in the Codex. Edges carry the relationship type. The graph persists across investigations, so the next question starts smarter than the last. Hover any node to highlight its neighbors. Drag to rearrange.
The shape of work that lands as a graded briefing. Concrete questions, scoped to the answer.
Who's winning the autonomous SOC race, and what's the installed-base moat worth through 2028?
How does Iran route value around sanctions through stablecoins, and what did the 2025 burn-wallet attack expose about the pipeline?
What's the M&A picture in AI-enabled cybersecurity, and which acquirers are most exposed to a competitor's pivot?
Is a US-Greenland compact for AI data-center build-out plausibly going through, and on what timeline?
Six stages between you and a graded answer. The autonomous parts run on the Codex graph. The human parts are non-negotiable.
You name a subject and a question. No scoping interview, no RFI dance.
Autonomous pre-scope of what the Codex already knows. ~90 seconds. Free.
Stripe, PayPal, or Bitcoin via BTCPay. Only if the preview is worth the spend.
Collection, deduplication, entity resolution, graph traversal, anomaly detection.
Grade. Adversarial pass. Human analyst review. No shortcuts at this stage.
Your report lands in your library. HTML, PDF, raw data export.
Annotated below: what you see on the cover of every Rook Report, and why it's there.
Most strategic intelligence still moves through five-figure subscriptions or six-week consulting cycles. Here's where the two paths diverge — line by line. Hover any row to see its counterpart.