voice-update — Claude Code skill: what it does, how to install
Speaks a short audio summary of completed tasks using pocket-tts. Useful when monitoring background runs without watching the screen.
npx skills add pchalasani/claude-code-tools --skill voice-update -a claude-codeWhat voice-update does
The skill instructs the agent to summarize finished tasks in one or two conversational sentences. It triggers speech via Bash using ${CLAUDE_PLUGIN_ROOT}/scripts/say, routing summary text into pocket-tts. The prompt explicitly drops code snippets, file paths, and technical jargon to keep the message natural.
You can specify an alternative voice using --voice (such as azure). When the pocket-tts background process is not running, the script launches it automatically. Expect the initial start to take roughly 30–60s.
Audio playback fails if your machine lacks uvx, along with afplay on macOS or aplay on Linux.
What it helps with
- hear task completion summaries out loud
- trigger spoken feedback when prompted by a Stop hook
- track agent progress without switching back to the terminal
Example prompts for the agent
- Give me a quick spoken summary of what you fixed
- Read out the final status once the test suite finishes
How to install voice-update
1Skills CLI (Vercel)
npx skills add pchalasani/claude-code-tools --skill voice-update -a claude-code2Manually: put the files in place
~/.claude/skills/voice-update/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/voice-update/
1Skills CLI (Vercel)
npx skills add pchalasani/claude-code-tools --skill voice-update -a codex2Manually: put the files in place
.agents/skills/voice-update/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add pchalasani/claude-code-tools --skill voice-update -a cursor2Manually: put the files in place
.cursor/skills/voice-update/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install pchalasani/claude-code-tools voice-update2Manually: put the files in place
.github/skills/voice-update/SKILL.md1Skills CLI (Vercel)
npx skills add pchalasani/claude-code-tools --skill voice-update -a gemini-cli2Platform CLI
gemini skills install https://github.com/pchalasani/claude-code-tools --path plugins/voice/skills/voice-updateManually: put the files in place
.devin/skills/voice-update/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add pchalasani/claude-code-tools --skill voice-update -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about voice-update
How to install voice-update in Claude Code?
Run npx skills add pchalasani/claude-code-tools --skill voice-update -a claude-code in your terminal.
What does the voice-update skill do?
It turns recent task progress into a short spoken update using pocket-tts. The generated summary stays conversational and leaves out paths or code snippets.
What dependencies are needed for voice-update?
The setup requires uvx for pocket-tts. For actual audio playback, you need afplay on macOS or aplay on Linux.
Source text
SKILL.md · MIT
---
name: voice-update
description: This skill should be used when the agent needs to give a spoken voice update to the user, or when reminded by a Stop hook to provide audio feedback. Use this skill to speak a short summary of what was accomplished.
---
# Voice Update Skill
Provide spoken audio feedback to the user using pocket-tts.
## When to Use
- When finishing a task and a Stop hook reminds to give voice feedback
- When the user explicitly asks for a spoken summary
- When providing important status updates that benefit from audio
## How to Use
1. Summarize what was accomplished in 1-2 short, conversational sentences
2. Call the `say` script with the summary text
## Calling the Say Script
Use Bash to call the say script:
```bash
${CLAUDE_PLUGIN_ROOT}/scripts/say "Your summary here"
```
Example:
```bash
${CLAUDE_PLUGIN_ROOT}/scripts/say "I've fixed the bug in the login handler and added the unit tests."
```
With a specific voice:
```bash
${CLAUDE_PLUGIN_ROOT}/scripts/say --voice azure "Task completed successfully."
```
## Summary Guidelines
- Keep it to 1-2 sentences maximum
- Be conversational, not robotic
- Match the user's communication style - if they're casual or use colorful language, mirror that tone
- Focus on what was accomplished, not technical details
- Avoid code snippets, file paths, or technical jargon
- Examples:
- "I've updated the configuration file and restarted the server."
- "The tests are now passing. I fixed three type errors."
- "Done! I created the new component and added it to the main page."
## Notes
- The say script auto-starts the pocket-tts server if not running (first use may take ~30-60s)
- Requires `uvx` and `afplay` (macOS) or `aplay` (Linux)
© pchalasani · MIT · pchalasani/claude-code-tools