Loading observability overview
Loading observability overview
Loading eval lab
Automated guided trials from npm run eval:run. Open an experiment to review per-case results and transcripts.
Experiments
0
Case pass rate
—
0/0 cases
Cases run
0
in selected period
p50 latency
—
min. 3 trials with latency
Avg latency
—
Avg validates
—
Avg user turns
—
Over cap
0
tuning signals
No automated experiments in this period. Run the guided eval harness locally to populate eval_trials.
npm run eval:run -- --case=static-browse-v0.2
Grouped model × prompt matrix comparison ships in Phase 7.7 (npm run eval:matrix).