Second revision: MCP servers, shared-link AI, flashcard modes, document previews
All checks were successful
Verify and publish V2 container / build-and-push (push) Successful in 30s

Generation
- Every quiz is NCLEX style; "Practice exam" writes 50 questions.
- Calculation questions only when the material involves calculations (max 5 per exam).
- Broad topics: the class's topics are listed in the prompt, a quick grouping call maps
  narrow labels onto broad ones, and leftover variants are merged.
- Exact select-all count per batch; option order shuffled.
- Python check: at most 3 runs, fixed problems count as passed, cosmetic remarks ignored.

Assistant
- Text, tool steps and cards appear in the order they happen, with paragraph breaks
  between steps.
- Optional thinking box: fixed height, follows the newest text, can be scrolled back.
- "+" opens a material picker listing recent items; items can be dragged in from class
  pages; shorter placeholder.

MCP servers
- Streamable HTTP client; encrypted headers.
- Per-server instructions and switches for chat, Explain, generation and the quiz check.
- Settings UI.

Share links
- Optional "Explain this" and "Ask about this material" per link, off by default.
- Items are loaded on the server and checked against the share; no tools; rate-limited.
- Migration 4 stores the switches.

Study and documents
- Standard flashcard mode (optionally repeat missed cards until known) next to spaced
  repetition.
- Streak visuals restyled in the app palette.
- Side by side / Original document views, with sandboxed previews of Word and PowerPoint
  files.

Other
- Built-in embedding model baked into the image; sharp stubbed if it cannot load; real
  load error shown in Settings.
- Python runner blocks network and process modules in JavaScript too.
- "View questions and answers" removed from the quiz start screen.
- Docs updated.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Elijah 2026-11-15 15:48:54 -07:00
parent 5ea61c882c
commit 031f7bec31
45 changed files with 2918 additions and 397 deletions

View file

@ -91,6 +91,11 @@ function toolArgs(name: string, userText: string, all: string): Record<string, u
const fix = /Q2 \[single\]/.test(all) ? [{ question: 2, action: 'set_correct', correct_options: [2], reason: 'mock: option 2 is the right answer' }] : [];
return { summary: 'Checked by the mock.', checks: [{ name: 'Question count', passed: true, detail: 'matches the request' }, { name: 'Correct answers per type', passed: true }], fixes: fix };
}
case 'submit_topics': {
// Labels of three or more words are broadened to their first word, like a real model would.
const labels = [...(userText.split('Labels:')[1] ?? '').matchAll(/^- (.+)$/gm)].map((m) => m[1].trim());
return { topics: labels.map((label) => ({ label, topic: label.split(/\s+/).length >= 3 ? label.split(/\s+/)[0] : label })) };
}
case 'get_performance':
case 'list_classes':
case 'list_materials':
@ -100,12 +105,14 @@ function toolArgs(name: string, userText: string, all: string): Record<string, u
return { reason: 'Requested by the user.', changes: [{ op: m?.[1]?.toLowerCase() ?? 'delete', kind: m?.[2]?.toLowerCase() ?? 'quiz', id: m?.[3] ?? 'missing', title: 'Renamed by mock' }] };
}
default:
return {};
return name.startsWith('mcp_') ? { drug: 'furosemide' } : {};
}
}
function chooseTool(userText: string, names: string[]): string | null {
const pick = (name: string) => (names.includes(name) ? name : null);
// "mcp" in the message: use the first tool from an MCP server.
if (/\bmcp\b/i.test(userText)) return names.find((n) => n.startsWith('mcp_')) ?? null;
if (/\b(delete|rename)\b/i.test(userText)) return pick('propose_changes');
if (/flashcard/i.test(userText)) return pick('generate_flashcards');
if (/\bquiz\b|practice questions/i.test(userText)) return pick('generate_quiz');
@ -200,6 +207,12 @@ export function startMockAi(port = 0): Promise<MockAi> {
return callTool(forced === 'submit_review' || ranPython ? 'submit_review' : 'run_python');
}
if (forced) return callTool(forced);
// "Some tool" (generation with MCP tools offered): an MCP tool first when asked for, then the real one.
if (body?.tool_choice === 'required' && names.length) {
const mcpTool = names.find((n) => n.startsWith('mcp_'));
if (last.role !== 'tool' && mcpTool && /\bmcp\b/i.test(userText)) return callTool(mcpTool);
return callTool(names.find((n) => !n.startsWith('mcp_')) ?? names[0]);
}
if (last.role === 'tool') {
const ref = /"ref":"(c\d+)"/.exec(text(last.content))?.[1];
if (/draft_ready|revisions_ready/.test(text(last.content))) return say('The draft is ready. Review it and save it when you are happy with it.');