add-voice
An agent skill by cursor, from cursor/plugins. Tags: audio, generative, integration, ui, voice.
What it does
Use when the user runs /add-voice, types Voice Mode, or asks to add Grok realtime voice to an app, including replacing an STT-LLM-TTS cascade or OpenAI Realtime. Wire speech-to-speech, safe auth, and app mic. Composer: waveform button, mic icon reserved for dictation. For mic-to-text only use /add-dictation; to speak text replies use /add-read-aloud. To add debug logging and fix from logs use /debug-voice.
Install
With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:
npx skills add cursor/plugins --skill add-voiceOr copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):
git clone --depth 1 https://github.com/cursor/plugins
cp -r plugins/grok-voice/skills/add-voice ~/.claude/skills/add-voiceSafety box score
Grade B, low risk: 21 of 100 risk points, rated 2026-10-20. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.
This skill wires an app to Grok realtime voice over a WebSocket and explicitly handles microphone audio, playback, and API authentication. The main risks are network exposure of live speech/audio to x.ai and the handling of long-lived API keys or ephemeral client secrets; it also adds a dev-only log sink inside the project.
| Category | Level | Why |
|---|---|---|
| Code & shell execution | 0 (none) | No native scripts or command execution are shipped in the skill; it is a markdown guide with code snippets only. |
| Network access | 2 (moderate) | It connects to `wss://api.x.ai/v1/realtime?model=grok-voice-latest` and `POST https://api.x.ai/v1/realtime/client_secrets`, so live audio and related session data leave the machine as part of the skill's purpose. |
| Filesystem reach | 1 (low) | The debug flow writes project-local logs to `.voice-logs/<sessionId>.ndjson` and expects app code changes, but nothing indicates access outside the working tree. |
| Secrets & credentials | 2 (moderate) | It uses `XAI_API_KEY` for server auth and ephemeral client secrets for browser/mobile, and explicitly warns not to put long-lived keys in client bundles. |
| Privilege & persistence | 0 (none) | There is no sudo, service installation, shell startup editing, or other privilege escalation in the skill. |
| Instruction hijack surface | 0 (none) | The file contains no hidden instructions, prompt-tricks, or directives to ignore safety rules; its instructions are operational, not adversarial. |
| Irreversible actions | 0 (none) | It does not push, delete, pay, book, or otherwise act externally on the user's behalf; it only configures voice I/O within an app. |
| Transparency | 0 (none) | The skill is plain text with readable snippets and links; there are no binaries, base64 blobs, or downloaded installers. |
Before installing
- Confirm you are willing to send live microphone audio to `api.x.ai` during voice sessions.
- Verify your browser/mobile path uses ephemeral client secrets, not a baked-in `XAI_API_KEY`.
- Read or implement the `/api/voice/log` path before enabling debug mode; it writes `.voice-logs/<sessionId>.ndjson`.
- Search the built client bundle for `XAI_API_KEY` and `client_secrets` to ensure no secret names or values are shipped.
Rated 2026-10-20 by an automated review of 1 file from the skill's repository.
Source
- Repository
- cursor/plugins (all skills from this repository)
- Path
- grok-voice/skills/add-voice/SKILL.md
- Branch
- main
- Collection
- cursor
- Updated
- 2026-09-23
Related skills
- add-read-aloud — Use when the user runs /add-read-aloud or wants the app to speak text with Grok text-to-speech: read-aloud button on assistant replies, auto-speak, TTS.
- add-dictation — Use when the user runs /add-dictation or wants speech turned into text with Grok speech-to-text: a mic button that dictates into the composer, live captions.
- book-sft-pipeline — This skill should be used for book-to-SFT pipelines: ePub extraction, literary segmentation, author-voice dataset construction, style-transfer training.
- music-to-video — Turn a music track (an audio file, a video to pull audio from, or a track generated from a mood brief) into a beat-synced video — lyric video, slideshow.
- setup-vivox-voice-chat — Add and configure in-game voice chat and text chat for Unity multiplayer games using Unity Vivox.
- stepfun-tts — Generate Chinese / Japanese speech with StepFun's Contextual TTS — default `stepaudio-2.5-tts` (blind-judged better on neutral/emotive preset voices).