System index
MICA measures systems, not models
An entry is a dated snapshot of a whole system: orchestrator, models, tools and memory together. No global model is required: a system may route each task to whichever model it judges best and call several inside one attempt, as long as every invocation is disclosed. Two entries can share a base model and still land far apart, which is the point.
Preview 0.1
6 markets · 10 task families · 3 separate outcome axes.
System registry
The registry is empty
No system snapshot has been submitted, verified and published. MICA leaves the page empty rather than fill it with illustrative entries: an invented name beside an invented figure is the exact failure a benchmark exists to prevent. The labels below define what an entry will carry when the first one is admitted.
Reading the labels
Verification status
Only independently rerun results are ever on the publication track. Everything in this demo edition is excluded regardless, because demo data can never be publication eligible.
- Independent rerun · IND
- MICA re-executed the submitted system snapshot on MICA-controlled accounts and infrastructure, and holds the full evidence trace.
- Provisional · PROV
- MICA observed a partial or supervised run, or holds evidence for only part of the claimed coverage. Reported, never published as official.
- Self-reported · SELF
- Submitted by the system's operator with an evidence trace MICA has not reproduced. Displayed for transparency only.
Reading the labels
Result tracks
- Simulator
- Deterministic local replicas of common local service flows. Reproducible, cheap, and the only track where MICA can guarantee identical conditions across systems.
- Live shadow
- The system acts against real services but stops at the confirmation boundary; no irreversible action is taken. Used to check that simulator behaviour transfers.
- Limited verified live
- A small number of end-to-end runs on MICA-held test accounts where the operator of the service has been notified. Narrow by design.