demo
Last 30 days · Jun 21 – Jul 21, 2026
1 Critical finding
13
Runs (30d)
46%
Success rate
1
Drift signals
$0.0553
Avg cost / run
Explore this project
Top signals
writer Agent behaviour latency_s, cost_usd, success
Versions
Version snapshots
Baseline is v1 — drift and comparisons are measured against it.
v1 Version 1 (baseline)
0
runs
0%
success
$0.0000
avg cost
Baseline
v2 v2 — analyst prompt · 2026-06-01
0
runs
0%
success
$0.0000
avg cost
v3 v3 — researcher web_search tool · 2026-06-11
3
runs
100%
success
$0.0130
avg cost
v4 v4 — writer gpt-4-turbo + long-form prompt · 2026-06-22
10
runs
30%
success
$0.0680
avg cost