Executive Summary
Week of Jul 6, 2026 — Jul 12, 2026
RouteIQ monitored 1 systems across your AI agent fleet, processing 15 runs at a total cost of $0.05. The fleet achieved a 0% success rate (15 bad runs caught before they escalated).
Success Rate
0%
Bad Runs
15
Cost This Week
$0.05
Open P1s
0
Why-Engine Track Record
76.6%
hypothesis accuracy
Of 47 automated root-cause hypotheses generated this period, 36 were subsequently confirmed by human review or metric recovery. This 76.6% confirmation rate is the trust signal behind every “[HYPOTHESIS]” annotation in your alert feed — when we say “model swap caused this”, we're right 76.6% of the time.
SLO Compliance
| System | Success Rate | Target | Status |
|---|---|---|---|
| sim-ecommerce | 0.0% | 95% | ✗ Fail |
Incidents
Cost
$0.05
this week
$0.20
projected / month
Top models by spend
| Model | Cost | Runs |
|---|---|---|
| gpt-4o | $0.02 | 9 |
| claude-sonnet-4-5 | $0.02 | 5 |
| gpt-4o-mini | $0.00 | 1 |
gpt-4o switch briefly cut per-run cost, but drove the quality failure that triggered the P1 incident — cheaper isn't always better when output quality is the SLO.Interventions & Audit
Week-over-Week
This week: 0% success across 15 runs at $0.05 total cost. Prior-week delta requires two consecutive report periods — available from week 2 onward.