Add knowledgebases

Knowledgebases give bots RAG over documents, scoped like extensions: to
particular bots, to a workspace, or to every bot, with their own button
beside Tools in the composer. Documents are chunked and embedded by the
file-ingest worker with the install's embedding model, which is now an
install-wide setting rather than a per-bot one.

- Word search runs on an FTS5 index ranked by BM25, with a scored LIKE
  fallback for scripts written without spaces.
- Changing the embedding model re-indexes documents automatically.
- Ingest jobs wait without spending attempts while every embedding model
  rests, fail at once on errors retrying can't fix, and keep passages
  searchable by words when the model refuses.
- Settings > Knowledgebases shows each knowledgebase's and document's
  index state: spinner, warning, grey check (words only) or green check.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

src/file-ingest/defer-ingest-job.ts

@@ -0,0 +1,27 @@
+import BunSQLite from "@moduletrace/bun-sqlite";
+import { AppData } from "@/src/data/app-data";
+
+type Params = {
+    job_id: number | string;
+    /**
+     * Unix ms before which the job can't be claimed again.
+     */
+    until: number;
+};
+
+/**
+ * Puts a claimed ingest job back to wait, giving back the attempt it
+ * used: every embedding model was resting, so nothing was tried. Capped
+ * at `AppData.HarnessMaxCooldownMs` from now.
+ */
+export default async function deferIngestJob({ job_id, until }: Params) {
+    const now = Date.now();
+    const at = Math.min(Math.max(until, now + AppData.HarnessCooldownMs), now + AppData.HarnessMaxCooldownMs);
+
+    await BunSQLite.sql({
+        sql: `UPDATE file_ingest_queue SET status = 'pending', locked_at = NULL, locked_by = NULL,
+            attempts = MAX(attempts - 1, 0), available_at = ?, updated_at = ?
+            WHERE id = ? AND status = 'processing'`,
+        values: [at, now, job_id],
+    });
+}