Source: ${html(e.source?.sha ?? result?.source?.sha ?? "Unknown")} · ${html(e.source?.ref ?? result?.source?.ref ?? "Unknown")} ${e.source?.dirty ? "(working tree modified; compare instruction hashes)" : ""}
${e.runtimeSettings?.onboardingRuntime ? `Setup: ${html((e.runtimeSettings.onboardingRuntime as { mode: string }).mode)}.
` : ""}Configured model: ${html(e.configuredModel ?? "Provider default")}. Observed: ${html(e.observedModels.join(", ") || "Not reported")}.
${html(JSON.stringify(e.runtimeSettings ?? {}, null, 2))}Source SHA-256: ${html(i.sha256)}
Credential redaction applied to this display copy. Display SHA-256: ${html(i.contentSha256)}
${html(i.content)}Quality scores are informational and never change behavioral pass/fail.
${ q ? `Judge: ${html(q.config.model)} · ${html(q.status)} · estimated cost: ${html(q.estimatedCostUsd === null ? "unknown" : `$${q.estimatedCostUsd.toFixed(6)}`)} · reservation: $${q.reservedCostUsd.toFixed(6)} · ${html(q.inputTokens ?? "?")} input / ${html(q.outputTokens ?? "?")} output tokens.
| Dimension | Score / 5 | Reason and evidence |
|---|---|---|
| ${html(s.dimension)} | ${s.score} | ${html(s.rationale)} ${s.evidence.map((id) => `${html(id)}`).join(" ")} |
${html(JSON.stringify(q, null, 2))}Not judged. Run the explicit judge command to add scores.
` }${html(JSON.stringify(t, null, 2))}${(
last?.documents ?? []
)
.filter((d) => d.issueId === t.id)
.map(
(d) =>
`${html(d.body)}${html(JSON.stringify(c, null, 2))}