Fifth revision: share-link AI safety, sturdier generation, cleaner UI
All checks were successful
Verify and publish V2 container / build-and-push (push) Successful in 1m0s

- Malformed model JSON is sent back for repair, then retried; a failed job
  starts over once before it fails. Visitors can dismiss failed sets.
- Generated sets get a short distinct name and description; flashcards are
  atomic and have no hints; standard study shows known/learning counts.
- Share links: "In progress" list, finished sets join the normal lists with a
  "New" tag, plain progress steps, newest exam first.
- Share-link AI expires after a week, can be limited to the US or Arizona
  (DB-IP Lite or Cloudflare headers) and is locked by abuse protection
  (migration 8).
- The assistant never answers from general knowledge; MCP and web search
  tools load without a manual test.
- Plain-text copy, no question count in exam mode, filler text removed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Elijah 2026-11-16 09:17:50 -07:00
parent d8df5edf30
commit 85d5c9186d
46 changed files with 1242 additions and 283 deletions

View file

@ -495,7 +495,7 @@ function ProviderCard({ provider, models, adapter, onEdit, reload }: { provider:
/* ------------------------------ tasks ------------------------------ */
function TaskRow({ name, label, description, config, models, resolved, onChange }: { name: TaskName | 'default'; label: string; description: string; config: TaskConfig; models: AiModel[]; resolved?: string; onChange: (c: TaskConfig) => void }) {
function TaskRow({ name, label, description, config, models, resolved, onChange }: { name: TaskName | 'default'; label: string; description?: string; config: TaskConfig; models: AiModel[]; resolved?: string; onChange: (c: TaskConfig) => void }) {
const model = models.find((m) => m.id === config.modelId);
const hasOverrides = config.temperature !== null || config.reasoningEffort !== null || config.maxOutputTokens !== null;
const byProvider = useMemo(() => {
@ -508,7 +508,7 @@ function TaskRow({ name, label, description, config, models, resolved, onChange
<div className="grid items-center gap-x-4 gap-y-2 sm:grid-cols-[1fr_280px]">
<div>
<div className="text-sm font-semibold">{label}</div>
<div className="text-[13px] text-ink-2">{description}</div>
{description && <div className="text-[13px] text-ink-2">{description}</div>}
</div>
<Select value={config.modelId ?? ''} onChange={(e) => onChange({ ...config, modelId: e.target.value || null })} aria-label={`Model for ${label}`}>
<option value="">{name === 'default' ? 'First available model' : `Same as default${resolved ? ` (${resolved})` : ''}`}</option>
@ -638,7 +638,6 @@ function Usage() {
)}
</div>
)}
<p className="mt-2 text-xs text-ink-3">Costs are estimates from token counts and the list prices on each model (editable). Your provider's bill is the source of truth.</p>
</Section>
);
}
@ -721,20 +720,16 @@ export default function AiSettings() {
))}
</div>
)}
<p className="mt-2.5 text-xs text-ink-3">
API keys are encrypted on the server and never sent back to the browser. These connections use each provider's API billing. ChatGPT plan sign-in requires a separate approved integration for remotely hosted apps; it is not configured in this installation.
</p>
</Section>
{usable.length > 0 && (
<Section title="Which model does what">
<div className="card divide-y divide-line">
<TaskRow name="default" label="Default model" description="Used for everything that has no model of its own below." config={tasks.data.default} models={usable} onChange={(c) => saveTask('default', c)} />
<TaskRow name="default" label="Default model" config={tasks.data.default} models={usable} onChange={(c) => saveTask('default', c)} />
{(Object.keys(meta.data.tasks) as TaskName[]).map((t) => (
<TaskRow key={t} name={t} label={meta.data!.tasks[t].label} description={meta.data!.tasks[t].description} config={tasks.data![t]} models={usable} resolved={status.data?.resolved[t]?.label} onChange={(c) => saveTask(t, c)} />
))}
</div>
<p className="mt-2.5 text-xs text-ink-3">You can also pick a different model for a single draft from "More options" when generating.</p>
</Section>
)}