DeepTutor: Lifelong Personalized Tutoring. https://deeptutor.info/.
Python ai-agentsai-tutorclawbotcli-tooldeepresearchinteractive-learninglearningmulti-agent-systemsrag
"use client";
import { Check, X } from "lucide-react";
import { useTranslation } from "react-i18next";
import type { ObjectiveReport } from "@/lib/learning-api";
import { formatAbsolute, formatRelative, type Translate } from "./format";
/**
* The evidence behind one objective.
*
* The map answers "am I through the gate"; this answers "why" — which
* questions were asked, what the learner said, what the engine did with it,
* and when it comes back for review. The gate itself is shown as a bar with
* the threshold marked, because a hard gate is only legible if you can see how
* far away it is.
*/
export function ObjectiveDetail({
report,
zh,
}: {
report: ObjectiveReport;
zh: boolean;
}) {
const { t } = useTranslation();
const qualitative = report.gate === "qualitative";
const evidence = report.evidence ?? [];
const evidenceCount = report.evidence_count ?? evidence.length;
return (
<div className="mt-1 mb-2 ml-5 space-y-3 border-l border-[var(--border)] pl-3">
<GateBar report={report} />
{report.review && (
<Row label={t("Spaced review")}>
{report.review.due_at
? t("Due {{relative}} · {{absolute}}", {
relative: formatRelative(report.review.due_at, zh),
absolute: formatAbsolute(report.review.due_at, zh),
})
: t("Not scheduled")}
<span className="ml-2 text-[var(--muted-foreground)]">
{t("memory {{retrievability}}% · stability {{stability}} days", {
retrievability: Math.round(report.review.retrievability * 100),
stability: report.review.stability.toFixed(1),
})}
</span>
<span className="block text-[var(--muted-foreground)]">
{t("risk {{risk}}% · {{lapses}} lapses", {
risk: Math.round(report.review.forgetting_risk * 100),
lapses: report.review.lapse_count,
})}
</span>
{report.review.reason && (
<span className="block text-[var(--muted-foreground)]">
{report.review.reason}
</span>
)}
</Row>
)}
{qualitative && report.explanation && (
<Row label={t("Your explanation")}>
<span className="italic">{report.explanation}</span>
</Row>
)}
{report.attempts.length > 0 && (
<div>
<div className="text-xs text-[var(--muted-foreground)]">
{t("Attempts ({{correct}}/{{total}} correct)", {
correct: report.correct_count,
total: report.attempts.length,
})}
</div>
<ul className="mt-1 space-y-1.5">
{[...report.attempts].reverse().map((attempt, index) => (
<li
key={`${attempt.question_id}-${index}`}
className="flex gap-2 text-xs"
>
{attempt.is_correct ? (
<Check className="mt-0.5 h-3 w-3 shrink-0 text-green-500" />
) : (
<X className="mt-0.5 h-3 w-3 shrink-0 text-red-500" />
)}
<div className="min-w-0">
<div className="text-[var(--foreground)]">
{attempt.prompt || t("(prompt unavailable)")}
</div>
<div className="text-[var(--muted-foreground)]">
{t("You said: ")}
{attempt.answer || t("(blank)")}
<span className="ml-2">
{formatRelative(attempt.at, zh)}
</span>
</div>
</div>
</li>
))}
</ul>
</div>
)}
{report.errors.length > 0 && (
<Row label={t("Error diagnosis")}>
{report.errors.map((record) => (
<span key={record.id} className="mr-2">
{t(record.error_type)}
{record.retries > 0 &&
t(" · {{count}} retries", { count: record.retries })}
{record.status === "graduated" && t(" · cleared")}
</span>
))}
</Row>
)}
{report.explained && <Row label={t("Explained")}>{report.explained.summary}</Row>}
{!!report.required_visual_tasks?.length && <Row label={t("Independent visual evidence")}>{report.demonstrated_visual_tasks?.length ?? 0} / {report.required_visual_tasks.length}</Row>}
{!!report.required_visual_tasks?.some((task) => !report.demonstrated_visual_tasks?.includes(task)) && <p className="text-xs text-[var(--muted-foreground)]">{t("Independent visual evidence is still required for this objective.")}</p>}
{report.attempts.some((attempt) => attempt.independent === false) && <p className="text-xs text-[var(--muted-foreground)]">{t("Assisted practice is recorded separately from independent mastery.")}</p>}
{evidence.length > 0 && (
<div>
<div className="text-xs text-[var(--muted-foreground)]">
{t("Retention evidence ({{count}})", {
count: evidenceCount,
})}
</div>
<ul className="mt-1 space-y-1">
{evidence.map((evidence, index) => (
<li
key={`${evidence.timestamp}-${evidence.turn_id || index}`}
className="flex items-center justify-between gap-2 text-[11px] text-[var(--muted-foreground)]"
>
<span>
{t(evidence.result)} · {t(evidence.assessment_type)}
{evidence.visual_context?.task && ` · ${t("Visual practice")}${evidence.hints_used ? ` · ${t("Assisted")}` : ""}`}
{evidence.quality !== null &&
` · ${Math.round(evidence.quality * 100)}%`}
</span>
<span className="shrink-0">{formatRelative(evidence.timestamp, zh)}</span>
</li>
))}
</ul>
{evidenceCount > evidence.length && (
<p className="mt-1 text-[10px] text-[var(--muted-foreground)]">
{t("Showing the latest 20 events")}
</p>
)}
</div>
)}
{report.attempts.length === 0 && !report.explanation && (
<p className="text-xs text-[var(--muted-foreground)]">
{t(
"No attempts yet. Once you work through this waypoint, its questions, your answers, and the grading evidence will appear here.",
)}
</p>
)}
</div>
);
}
/** Mastery against the gate it has to clear. */
function GateBar({ report }: { report: ObjectiveReport }) {
const { t } = useTranslation();
const qualitative = report.gate === "qualitative";
// A qualitative gate is a boolean, so all-or-nothing is the only honest
// fill. `mastery` carries quiz accuracy for these objectives too, and
// drawing that as gate progress is how a waypoint comes to show a full bar
// beside a hollow dot — the learner reads "100%, still not cleared".
const pct = qualitative
? report.mastered
? 100
: 0
: Math.round(report.mastery * 100);
const thresholdPct = Math.round(report.threshold * 100);
const perfectButBelowGate =
report.gate === "quantitative" &&
report.attempts.length > 0 &&
report.correct_count === report.attempts.length &&
report.mastery < report.threshold;
return (
<div>
<div className="flex items-baseline justify-between text-xs">
<span className="text-[var(--muted-foreground)]">
{report.gate === "qualitative"
? t("Qualitative gate: explain it in your own words")
: t("Quantitative gate: {{pct}}%", { pct: thresholdPct })}
</span>
<span
className={
report.mastered
? "text-green-500"
: "text-[var(--muted-foreground)]"
}
>
{pct}%
</span>
</div>
<div className="relative mt-1 h-1.5 w-full overflow-hidden rounded-full bg-[var(--accent)]">
<div
className={`h-full ${report.mastered ? "bg-[var(--primary)]" : "bg-[var(--muted-foreground)]"}`}
style={{ width: `${pct}%` }}
/>
{report.gate === "quantitative" && (
<div
className="absolute inset-y-0 w-px bg-[var(--foreground)]/40"
style={{ left: `${thresholdPct}%` }}
/>
)}
</div>
{perfectButBelowGate && (
<p className="mt-1.5 text-[11px] leading-4 text-[var(--muted-foreground)]">
{t(
"Even with {{correct}}/{{total}} correct so far, mastery also weighs evidence volume, difficulty, and recent consistency. One more discriminating practice can raise it further.",
{ correct: report.correct_count, total: report.attempts.length },
)}
</p>
)}
{qualitative && !report.mastered && report.attempts.length > 0 && (
<p className="mt-1.5 text-[11px] leading-4 text-[var(--muted-foreground)]">
{t(
"Practice questions do not open this gate — it opens when you explain the idea in your own words and your tutor records that. Your answers here still count as practice.",
)}
</p>
)}
</div>
);
}
function Row({
label,
children,
}: {
label: string;
children: React.ReactNode;
}) {
return (
<div className="text-xs">
<div className="text-[var(--muted-foreground)]">{label}</div>
<div className="mt-0.5 text-[var(--foreground)]">{children}</div>
</div>
);
}