refactor(open-sse): registry consolidation + DRY media/oauth/adhoc cleanup

- Single-source registry: oauth clientId/tokenUrl, usage URLs, image/embed
  configs, search defaultModel, codex fixedPort, google token url derive.
- Remove 29 unused OmniRoute providers (registry 100→71); media intact.
- De-adhoc: codex literals → registry format/oauth flags; reasoningInject,
  image/embed openrouter headers + xai bodyFields config-driven.
- Add REGISTRY_TEMPLATE.js + expand PROVIDER_DEFAULTS/schema JSDoc.
- Baselines updated; PROVIDERS 62 + alias 90 byte-for-byte, golden snapshots.

Co-authored-by: Cursor <cursoragent@cursor.com>

open-sse/handlers/ttsProviders/gemini.js

@@ -1,9 +1,10 @@
 // Gemini TTS — generateContent with AUDIO modality returns PCM L16, wrap as WAV
 import { Buffer } from "node:buffer";
+import { PROVIDER_MEDIA } from "../../providers/index.js";
 
-const DEFAULT_MODEL = "gemini-2.5-flash-preview-tts";
+const KNOWN_MODELS = (PROVIDER_MEDIA["gemini"]?.ttsConfig?.models || []).map((m) => m.id);
+const DEFAULT_MODEL = KNOWN_MODELS[0];
 const DEFAULT_VOICE = "Kore";
-const KNOWN_MODELS = ["gemini-2.5-flash-preview-tts", "gemini-2.5-pro-preview-tts"];
 
 // Parse "model/voice" — if input doesn't match a known TTS model, treat it as voice with default model
 function parseGeminiModelVoice(input) {