{"skill":{"slug":"discord-voice-deepgram","displayName":"Discord Voice Using Deepgram","summary":"Voice-channel conversations in Discord using Deepgram streaming STT + low-latency TTS","description":"---\nname: deepgram-discord-voice\ndescription: Voice-channel conversations in Discord using Deepgram streaming STT + low-latency TTS\nmetadata:\n  clawdbot:\n    config:\n      requiredEnv:\n        - DISCORD_TOKEN\n        - DEEPGRAM_API_KEY\n      optionalEnv: []\n      example: |\n        {\n          \"plugins\": {\n            \"entries\": {\n              \"deepgram-discord-voice\": {\n                \"enabled\": true,\n                \"config\": {\n                  \"streamingSTT\": true,\n                  \"streamingTTS\": true,\n                  \"ttsVoice\": \"aura-2-thalia-en\",\n                  \"vadSensitivity\": \"medium\",\n                  \"bargeIn\": true,\n\n                  \"primaryUser\": \"atechy\",\n                  \"allowVoiceSwitch\": true,\n                  \"wakeWord\": \"openclaw\",\n\n                  \"deepgram\": {\n                    \"sttModel\": \"nova-2\",\n                    \"language\": \"en-US\"\n                  }\n                }\n              }\n            }\n          }\n        }\n---\n\n# Deepgram Discord Voice (Clawdbot/OpenClaw Plugin)\n\nThis plugin lets you talk to your agent **only from a Discord voice channel**.\n\nPipeline (low latency):\n- Discord voice audio → **Deepgram streaming STT** (WebSocket)\n- Transcript → your agent\n- Agent reply → **Deepgram TTS** (`/v1/speak` streamed HTTP Ogg/Opus)\n- Audio played back into the voice channel\n\n## Requirements\n\n- A Discord bot token (`DISCORD_TOKEN`)\n- A Deepgram API key (`DEEPGRAM_API_KEY`)\n- Discord bot permissions in your server:\n  - **Connect**\n  - **Speak**\n  - **Use Voice Activity**\n\n## Install\n\n### Option A: Install from ClawHub\n\n1. In your OpenClaw/Clawdbot dashboard, open **Skills/Plugins**.\n2. Add/install **deepgram-discord-voice**.\n3. Set the required environment variables.\n\n### Option B: Manual install\n\n1. Copy this folder into your extensions/plugins directory.\n2. Run:\n\n```bash\nnpm install\n```\n\n3. Restart OpenClaw/Clawdbot.\n\n## Configuration\n\n### Key settings\n\n- `primaryUser` (recommended): Who the bot listens to by default.\n  - Best: your **Discord user ID** (numeric)\n  - Also supported: username/display name (e.g., `atechy`) if unique in-channel\n\n- `allowVoiceSwitch`: If `true`, the primary user can switch who is allowed by voice.\n\n- `wakeWord`: Prefix for voice control commands. Default: `openclaw`.\n\n- `deepgram.sttModel`: Default `nova-2`.\n- `deepgram.language`: Optional BCP‑47 language tag (e.g., `en-US`, `es`, `es-EC`).\n- `ttsVoice`: Deepgram Aura voice model (e.g., `aura-2-thalia-en`).\n\n### Example config\n\n```json5\n{\n  \"plugins\": {\n    \"entries\": {\n      \"deepgram-discord-voice\": {\n        \"enabled\": true,\n        \"config\": {\n          \"streamingSTT\": true,\n          \"streamingTTS\": true,\n\n          \"primaryUser\": \"atechy\",\n          \"allowVoiceSwitch\": true,\n          \"wakeWord\": \"openclaw\",\n\n          \"ttsVoice\": \"aura-2-thalia-en\",\n          \"vadSensitivity\": \"medium\",\n          \"bargeIn\": true,\n\n          \"deepgram\": {\n            \"sttModel\": \"nova-2\",\n            \"language\": \"en-US\"\n          }\n        }\n      }\n    }\n  }\n}\n```\n\n## Usage\n\n### Join a voice channel\n\nUse the plugin tool or slash command (depends on your OpenClaw setup):\n- Join: `action=join` with the `channelId`\n- Leave: `action=leave`\n\n### Talk (voice channel)\n\nOnce the bot is connected, just speak.\n\n### Safeguard: only listen to you (default)\n\nWhen `primaryUser` is set, the plugin will only listen to that user unless you allow someone else.\n\n### Let someone else talk (voice commands)\n\nAs the primary user, say:\n- `openclaw allow <name>`\n- `openclaw listen to <name>`\n\nTo lock it back:\n- `openclaw only me`\n- `openclaw reset`\n\n### Switch via tool actions (optional)\n\n- `allow_speaker` with `user` (id / @mention / name)\n- `only_me`\n- `status`\n\n## Notes\n\n- Lowest latency comes from `streamingSTT=true` and `streamingTTS=true`.\n- Deepgram TTS is streamed over HTTP in **Ogg/Opus** so Discord can play it immediately.\n","tags":{"latest":"1.0.0"},"stats":{"comments":2,"downloads":2022,"installsAllTime":1,"installsCurrent":1,"stars":5,"versions":1},"createdAt":1770770145129,"updatedAt":1779076922520},"latestVersion":{"version":"1.0.0","createdAt":1770770145129,"changelog":"Initial release of Deepgram Discord Voice skill for Clawdbot/OpenClaw.\n\n- Enables real-time voice conversations in Discord voice channels using Deepgram streaming STT and low-latency TTS.\n- Configurable to listen only to a specified primary user, with optional voice-activated speaker switching.\n- Supports wake-word activated voice commands for controlling permissions.\n- Provides example configurations and clear setup instructions.\n- Requires DISCORD_TOKEN and DEEPGRAM_API_KEY environment variables.\n- Delivers agent replies via Deepgram Aura TTS, streamed directly to the channel for minimal latency.","license":null},"metadata":{"setup":[],"os":null,"systems":null},"owner":{"handle":"adriel1006","userId":"s17ddkj6m6vw1280m6403h9vm1884fbc","displayName":"adriel1006","image":"https://avatars.githubusercontent.com/u/19422850?v=4"},"moderation":{"isSuspicious":false,"isMalwareBlocked":false,"verdict":"clean","reasonCodes":["review.llm_review"],"summary":"Review: review.llm_review","engineVersion":"v2.4.24","updatedAt":1779968647151}}