First revision: DeepSeek, numeric questions, background assistant, UI fixes
All checks were successful
Verify and publish V2 container / build-and-push (push) Successful in 1m2s

Providers and pricing
- Add a DeepSeek provider: thinking mode with low/high/max effort, reasoning_content
  replayed between tool steps, cache-hit usage, /models discovery and catalog prices.
- Fix "does not support tool calling": a rejected forced tool_choice (DeepSeek thinking
  mode) now asks for the call in the prompt; only a genuine no-tools 400 is "unsupported".
- Seven reasoning-effort levels (Off to Max) mapped per adapter.
- Cached-input price per model, cache-write and DeepSeek off-peak aware cost estimates.
- Built-in bge-small embedding model (transformers.js) as the semantic search default.

Quizzes
- Numeric (type-the-answer) questions: units ignored, optional ± tolerance; editor,
  runner, importer, templates and generation prompts.
- Quiz start screen instead of showing answers; navigator ring no longer clipped and
  key legend removed; clock only for timed exams; 7-point letter grades.

Assistant and generation
- Assistant turns run on the server independently of the browser, with re-attach and
  an explicit stop; drafts tray shows generations in progress.
- Per-chat model switcher, @-tagging of class materials, full-width assistant page and
  a smaller assistant button.
- Editable generation instructions and built-in presets, separate from the chatbot
  import instructions.

Other
- Dialog footer padding, keyboard hints removed, copy buttons work over plain HTTP,
  class and folder shares include documents.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Elijah 2026-11-15 12:48:17 -07:00
parent ccdad9728b
commit f0aabe4d5d
63 changed files with 3892 additions and 488 deletions

View file

@ -4,7 +4,7 @@
* connection-test wrapper.
*/
import { jsonrepair } from 'jsonrepair';
import { AIError, type DiscoveredModel, type ProviderKind, type TestResult, type Usage } from '../types.ts';
import { AIError, REASONING_EFFORTS, type DiscoveredModel, type ProviderKind, type ReasoningEffort, type TestResult, type Usage } from '../types.ts';
import { lookupModelDefaults } from './catalog.ts';
import { isRecord } from './http.ts';
@ -151,3 +151,22 @@ export async function runTest(label: string, check: () => Promise<{ modelCount?:
/** Timeout used by `test()` so a dead endpoint fails fast instead of hanging the settings page. */
export const TEST_TIMEOUT_MS = 20_000;
/**
* Maps a requested effort onto the closest level a provider accepts. Ties go to the higher
* level, so "medium" on a low/high-only API becomes "high" rather than silently weaker.
*/
export function nearestEffort<T extends string>(effort: ReasoningEffort, allowed: readonly T[]): T {
const rank = (e: string) => REASONING_EFFORTS.indexOf(e as ReasoningEffort);
const want = rank(effort);
let best = allowed[0];
let bestDistance = Number.POSITIVE_INFINITY;
for (const level of allowed) {
const d = Math.abs(rank(level) - want);
if (d < bestDistance || (d === bestDistance && rank(level) > rank(best))) {
best = level;
bestDistance = d;
}
}
return best;
}