Regression
0
Skip to workspace content
Org status
No active org
Sign in to activate an organization workspace.
AI Reliability
AI Eval Artifact Dashboard
Read-only visibility over local AI eval artifacts, citation freshness aging, trend history, and active warnings.
Source directory: /var/task/artifacts/evals
Severity policy: /var/task/config/ai-eval-regression-policy.json
No local eval artifacts yet
Run pnpm ai:eval:gate or pnpm ai:eval:report to generate artifacts/evals/*. The dashboard stays non-failing when ignored local artifacts are absent.
Severity trend distribution
Last 0 of configured 10 latest-vs-previous comparisons by review priority.
Watch
0
Stable
0
Improved
0
Severity sparkline
Window: 0/10 retained comparisons.
Need at least two retained runs before severity history can be shown.
Latest vs previous run
Need at least two retained eval runs before a latest-vs-previous diff can be shown.
What changed notes
Diff notes become available after two retained eval artifacts exist.
Artifact timestamps
- latestmissing ·
/var/task/artifacts/evals/ai-eval-gate-latest.json - trend-indexmissing ·
/var/task/artifacts/evals/trend-index.json - summarymissing ·
/var/task/artifacts/evals/summary.md
Workflow docs: AI EvalOps guide