Each dot is one attack the engine found. The picture shows the trade-off between how expensive the attack was and how much damage it caused. Use it to spot the cheap-but-nasty attacks — the bottom-right corner.
How to read it
- X axis — attack cost (lower is cheaper for the attacker).
- Y axis — severity (higher means a worse outcome for the system).
- Only Pareto-optimal points are shown: each dot is the best severity reachable at that cost.
- Bottom-right dots are the most concerning — large damage for small effort.
How to use it
- Pick a preset, or click Choose pareto_front.json… to load your own.
- Hover a dot to see its archive index, cost, and severity.
- Click a dot to pin it; use ← → to step through the archive.
Bundled samples cover the aggregate, network, resource-cascade, service-backlog, and liquidity-ladder domains, the coupled research fork, and the flagship preset. Related views: Replay viewer (play back one attack) · Attribution viewer (why an attack worked) · Composite viewer.
Load JSON to begin.