Loops (30d)
—
Iterations saved
—
Rollbacks
—
Outcomes
—
Loop Health Map
Most recent loops, color-coded by outcome. Earliest top-left, most recent bottom-right.
converged
high gain (≥ 0.85)
oscillating
diverged
max iterations
Convergence Profiles
Smoothed loop-gain (median Aβ per loop) over time. Drift upward = prompts or models getting harder.
Waste Report
Iterations saved vs an assumed fixed cap of 10. Multiply by your per-iteration cost for $-value.
Iterations saved
—
Estimated savings
—
Per-loop average
—