Add Deepgram MCP Server - speech-to-text and TTS

Python FastMCP server wrapping Deepgram API for audio transcription
and text-to-speech. Supports 125+ multilingual voices, large file
chunking via FFmpeg, formatted markdown output with speaker
diarization, and Docker deployment on port 8009.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

deepgram-mcp/docker-compose.yml

@@ -0,0 +1,21 @@
+services:
+  deepgram-mcp:
+    build: .
+    container_name: deepgram-mcp
+    restart: unless-stopped
+    ports:
+      - "8009:8009"
+    volumes:
+      - deepgram-uploads:/data/uploads
+      - deepgram-tts:/data/tts_output
+    env_file:
+      - .env
+    environment:
+      - UPLOAD_DIR=/data/uploads
+      - TTS_DIR=/data/tts_output
+      - HOST=0.0.0.0
+      - PORT=8009
+
+volumes:
+  deepgram-uploads:
+  deepgram-tts: