Evaluations
Continuous quality signal for the agentic FX — 4 layers, golden-set scenarios, trends, regressions.
Loading…
Pass rate by layer
Domain5/5 · 100%
Skill4/4 · 100%
Agent6/6 · 100%
Pipeline1/1 · 100%
Pass rate (30 days)
Loading…Cost per run (30 days)
Loading…