Execute everything.
No governor. Every proposed candidate is executed before the next decision.
One lane executes every candidate. The other asks MARGINAL before the spend. Same verifier. Same target. You control the clock.
No governor. Every proposed candidate is executed before the next decision.
Press RUN. MARGINAL will score each candidate before spend.
This is the deterministic result of this demo fixture, not a production savings claim.
Allocation decisions are replayed from the generated result data.
Deterministic functional demonstration using declared action-cost estimates; not provider telemetry, not a production benchmark, and not a claim about every agent workload.
Playback timing is accelerated for the browser. Declared costs and latency are demo inputs, not live provider billing. Build agents that spend compute deliberately.
marginal killer-demo --output killer-demo-output