feat: voice dictation (STT wave 1 — Deepgram cloud)

Two-band architecture from docs/research/stt.md: renderer captures mic
audio in an AudioWorklet (linear16 16kHz mono, ~128ms frames, per-quantum
RMS for the meter) and streams it directly to Deepgram over WebSocket —
audio never loops through our server. The real key stays server-side;
/api/stt/token mints a short-lived grant token (TTL 300s) per run.

Patterns ported from the owner's pi dictate extension (tuned in real
use): generation counter guarding every async handler, the
latest-RMS-never-resets meter (-50..-10dB window, 60ms tick), mic-first
stop + CloseStream + 3s safety net, endpointing=300 with the
utterance_end_ms=1000 backstop.

UI (owner placement): mic button in the stats bar's middle slot on all
three surfaces (home graph-view, chat, draft); while recording the meter
unfurls to fill the row's free width with the interim ghost at its
flank. Esc ends dictation from the composer. Token mint comes before
getUserMedia so a keyless app fails fast without a TCC prompt; one
automatic reconnect on abnormal socket close, errors surface inline.

Live-smoked the full error path in the dev shell (bar renders on all
three pages, 403 grant failure shows inline, mic untouched). The full
audio path needs a Deepgram key with Member permission to mint tokens.

components/home.tsx

@@ -5,6 +5,7 @@ import { useRouter } from "next/navigation";
 import { CornerDownLeftIcon } from "lucide-react";
 import { GraphView } from "@/components/graph-view";
 import { GraphContextControl, effectiveContextTokens } from "@/components/graph-context-control";
+import { DictationBar, useDictation } from "@/components/dictation";
 import { SidebarToggle } from "@/components/sidebar";
 import { StatsBar } from "@/components/stats-bar";
 import type { ChatSettings, TurnUsage } from "@/src/core";
@@ -18,6 +19,11 @@ import type { ChatSettings, TurnUsage } from "@/src/core";
 export function Home() {
 	const router = useRouter();
 	const [topic, setTopic] = useState("");
+	// Voice dictation (STT wave, 2026-09-17): finals append at the end of the
+	// topic draft; the strip lives in the stats bar's middle slot.
+	const dictation = useDictation({
+		append: (t) => setTopic((v) => (v ? v.replace(/\s+$/, "") + " " + t : t)),
+	});
 	const [busy, setBusy] = useState(false);
 	// The model+thinking selection made pre-chat — lifted from the StatsBar's
 	// ModelControl, carried to the chat POST then PUT-on-create (the model
@@ -97,6 +103,12 @@ export function Home() {
 						value={topic}
 						onChange={(e) => setTopic(e.target.value)}
 						onKeyDown={(e) => {
+							// Esc ends dictation (the dictate decisions: click-again or Esc).
+							if (e.key === "Escape" && dictation.active) {
+								e.preventDefault();
+								dictation.stop();
+								return;
+							}
 							if (e.key === "Enter" && !e.shiftKey) {
 								e.preventDefault();
 								void start();
@@ -121,6 +133,7 @@ export function Home() {
 					    budget (session-data slot), while the RIGHT keeps the model +
 					    thinking control (deferred save). */}
 					<StatsBar
+						dictationSlot={<DictationBar dictation={dictation} />}
 						turns={[] as TurnUsage[]}
 						chatId=""
 						settings={settings}