feat(call): in-call soundboard, quality controls, room call-permissions

Element Call is now consumed as our self-built fork
(@lotusguild/element-call-embedded); wire up its previously-dormant
capabilities and document the fork as live.

Soundboard (P5-15): a call-bar button plays user-uploaded audio clips into the
call as a real published track (io.lotus.inject_audio) plus local playback.
Clips are uploadable like emoji/sticker packs, stored in io.lotus.soundboard
account data (synced across devices). Gated by a Settings toggle + volume.

Quality controls (P5-31): per-user mic/screenshare bitrate + screenshare
framerate (Settings -> Calls), applied via io.lotus.set_quality clamped to any
room cap. Room admins set caps and hard call-permissions (allow_screenshare /
allow_camera) in Room Settings -> Voice; the call bar hides blocked buttons.

- New: CallSoundboard, useSoundboard, soundboardClips; RoomQuality,
  useCallQuality, callQuality (+ unit tests).
- Optimistic-write RoomQuality admin UI (no stale-state clobber).
- Docs: mark EC fork live across README/FEATURES/TODO/BUGS/TESTING; add D2
  manual-test steps.

Numeric quality caps are client-cooperative; screenshare/camera permissions are
hard-enforced server-side (see LotusGuild/matrix voice-limit-guard).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

src/app/plugins/call/CallControl.ts

@@ -7,6 +7,17 @@ export enum CallControlEvent {
   StateUpdate = 'state_update',
 }
 
+/**
+ * [lotus #7 / P5-31] Payload for the fork's `io.lotus.set_quality` action.
+ * All fields optional; `null` clears that cap. Bits/sec for bitrates, fps for
+ * framerate.
+ */
+export type LotusQualityPayload = {
+  audioMaxBitrate?: number | null;
+  screenshareMaxBitrate?: number | null;
+  screenshareMaxFramerate?: number | null;
+};
+
 export class CallControl extends EventEmitter implements CallControlState {
   private state: CallControlState;
 
@@ -358,6 +369,33 @@ export class CallControl extends EventEmitter implements CallControlState {
     this.call.transport.send('io.lotus.focus_participant', { userId: null }).catch(() => undefined);
   }
 
+  /**
+   * [lotus #3 / P5-15] Inject a soundboard clip into the call so other
+   * participants hear it. The fork publishes it as a separate LiveKit audio
+   * track (`io.lotus.inject_audio`) rather than splicing the mic. `url` must be
+   * an https/blob URL the widget can fetch WITHOUT credentials — the host
+   * resolves an mxc clip to a `blob:` object URL first (authenticated media
+   * can't be fetched cross-realm by the widget). `volume` is 0–1.
+   *
+   * The local user does not hear their own published track, so callers should
+   * also play the clip locally for feedback.
+   */
+  public injectAudio(url: string, volume = 1): void {
+    this.call.transport.send('io.lotus.inject_audio', { url, volume }).catch(() => undefined);
+  }
+
+  /**
+   * [lotus #7 / P5-31] Apply audio/screenshare encoding limits to the local
+   * published tracks (the fork's `io.lotus.set_quality` action, via
+   * `RTCRtpSender.setParameters` — no republish). Bitrates are bits/sec,
+   * framerate is fps. A field set to `null` clears that cap. Settings are
+   * sticky fork-side (re-applied on every re-publish / reconnect). Values are
+   * clamped fork-side, so out-of-range input can't brick the encoder.
+   */
+  public setQuality(settings: LotusQualityPayload): void {
+    this.call.transport.send('io.lotus.set_quality', settings).catch(() => undefined);
+  }
+
   public dispose() {
     this.bodyMutationObserver.disconnect();
     this.controlMutationObserver.disconnect();