Evals
CI reporters
JUnit, Markdown, GitHub annotations — for any CI stack.
import {
reportToCi,
renderJUnit,
renderMarkdown,
renderGitHubAnnotations,
} from '@agentskit/eval'
const report = await runEval({ agent, suite })
await reportToCi({
report,
output: [
{ kind: 'junit', path: 'eval-report.xml' },
{ kind: 'markdown', path: 'eval-report.md' },
{ kind: 'github-annotations' }, // auto-writes to ::error / ::warning
],
})GitHub Actions
- run: pnpm eval
- uses: actions/upload-artifact@v4
with:
name: eval-report
path: eval-report.md