helio-prodDashboard
Dashboard
Is your LLM spend safe? Are your agents looping? Is TokenPolice paying for itself?
Sessions
7.8k+3.3%
Model Calls
23.2k-13.9%
Spend on Models
$1847.32+18.4%
Blocked Calls
300+100%
enforced
Savings if Enforced
$12.60
21 calls would-block · est.
Model Reroutes
2.6k
150 would-route
Allowed vs Blocked Calls
Stacked by source · enforce solid / dry-run hatched
Block rate
1.38%
300 enforce · 21 dry-run
Spend Trend
Solid: this period · Dashed: previous period
Total
$1847.32
23,184 req
Unit Economics
Per session · vs previous period
Cache hit+2.1%
18.4%
1816.1k input tokens served from cache
$ / session
7,842 sessions
-4.8%$0.24
Runs / session
agent runs per session
+0.4%2.0
Blended $ / 1K tokens
effective price across all models
-1.2%$0.15
Loop detection
89 firingsHash cycle
86
86 unique traces
Cap
0 (+ 3 shadow)
3 unique traces
Loopiest workflows
View allinbox-router
89 traces · 89 firings
86(+3)
$225.40 on looping traces